Enhance token refresh functionality across multiple executors

- Updated refreshCredentials methods in various executors (Antigravity, Base, Default, Github, Kiro) to accept optional proxyOptions for improved proxy handling. - Modified token refresh logic to utilize proxy-aware fetch for better network management. - Enhanced usage retrieval functions to support proxy options, ensuring seamless integration with proxy configurations. - Updated ModelSelectModal and ProviderInfoCard components to incorporate kind filtering for improved user experience in model selection. - Added validation for API keys in the provider validation route, including support for webSearch/webFetch providers.
2026-05-08 12:01:28 +00:00 · 2026-04-28 17:28:57 +07:00
parent 1bb621317d
commit 8f81363675
45 changed files with 2924 additions and 289 deletions
--- a/README.md
+++ b/README.md
@@ -1,9 +1,9 @@
 <div align="center">
  <img src="./images/9router.png?1" alt="9Router Dashboard" width="800"/>
  
-  # 9Router - Free AI Router
+  # 9Router - FREE AI Router & Token Saver
  
-  **Never stop coding. Auto-route to FREE & cheap AI models with smart fallback.**
+  **Never stop coding. Save 20-40% tokens with RTK + auto-fallback to FREE & cheap AI models.**
  
  **Connect All AI Code Tools (Claude Code, Cursor, Antigravity, Copilot, Codex, Gemini, OpenCode, Cline, OpenClaw...) to 40+ AI Providers & 100+ Models.**
  
@@ -20,19 +20,21 @@

 ## 🤔 Why 9Router?

-**Stop wasting money and hitting limits:**
+**Stop wasting money, tokens and hitting limits:**

 - ❌ Subscription quota expires unused every month
 - ❌ Rate limits stop you mid-coding
+- ❌ Tool outputs (git diff, grep, ls...) burn tokens fast
 - ❌ Expensive APIs ($20-50/month per provider)
 - ❌ Manual switching between providers

 **9Router solves this:**

+- ✅ **RTK Token Saver** - Auto-compress tool_result content, save 20-40% tokens per request
 - ✅ **Maximize subscriptions** - Track quota, use every bit before reset
 - ✅ **Auto fallback** - Subscription → Cheap → Free, zero downtime
 - ✅ **Multi-account** - Round-robin between accounts per provider
- ✅ **Universal** - Works with Claude Code, Codex, Gemini CLI, Cursor, Cline, any CLI tool
+- ✅ **Universal** - Works with Claude Code, Codex, Cursor, Cline, any CLI tool

 ---

@@ -40,25 +42,26 @@

 ```
 ┌─────────────┐
-│  Your CLI   │  (Claude Code, Codex, Gemini CLI, OpenClaw, Cursor, Cline...)
+│  Your CLI   │  (Claude Code, Codex, OpenClaw, Cursor, Cline...)
 │   Tool      │
 └──────┬──────┘
       │ http://localhost:20128/v1
       ↓
-┌─────────────────────────────────────────┐
-│           9Router (Smart Router)        │
-│  • Format translation (OpenAI ↔ Claude) │
-│  • Quota tracking                       │
-│  • Auto token refresh                   │
-└──────┬──────────────────────────────────┘
+┌─────────────────────────────────────────────┐
+│           9Router (Smart Router)            │
+│  • RTK Token Saver (cut tool_result tokens) │
+│  • Format translation (OpenAI ↔ Claude)     │
+│  • Quota tracking                           │
+│  • Auto token refresh                       │
+└──────┬──────────────────────────────────────┘
       │
-       ├─→ [Tier 1: SUBSCRIPTION] Claude Code, Codex, Gemini CLI
+       ├─→ [Tier 1: SUBSCRIPTION] Claude Code, Codex, GitHub Copilot
       │   ↓ quota exhausted
       ├─→ [Tier 2: CHEAP] GLM ($0.6/1M), MiniMax ($0.2/1M)
       │   ↓ budget limit
-       └─→ [Tier 3: FREE] iFlow, Qwen, Kiro (unlimited)
+       └─→ [Tier 3: FREE] Kiro, OpenCode Free, Vertex ($300 credits)

-Result: Never stop coding, minimal cost
+Result: Never stop coding, minimal cost + 20-40% token savings via RTK
 ```

 ---
@@ -76,15 +79,15 @@ npm install -g 9router

 **2. Connect a FREE provider (no signup needed):**

-Dashboard → Providers → Connect **Claude Code** or **Antigravity** → OAuth login → Done!
+Dashboard → Providers → Connect **Kiro AI** (free Claude unlimited) or **OpenCode Free** (no auth) → Done!

 **3. Use in your CLI tool:**

 ```
-Claude Code/Codex/Gemini CLI/OpenClaw/Cursor/Cline Settings:
+Claude Code/Codex/OpenClaw/Cursor/Cline Settings:
  Endpoint: http://localhost:20128/v1
  API Key: [copy from dashboard]
-  Model: if/kimi-k2-thinking
+  Model: kr/claude-sonnet-4.5
 ```

 **That's it!** Start coding with FREE AI models.
@@ -233,30 +236,27 @@ Default URLs:
 <div align="center">
  <table>
    <tr>
-      <td align="center" width="150">
-        <img src="./public/providers/iflow.png" width="70" alt="iFlow"/><br/>
-        <b>iFlow AI</b><br/>
-        <sub>8+ models • Unlimited</sub>
-      </td>
-      <td align="center" width="150">
-        <img src="./public/providers/qwen.png" width="70" alt="Qwen"/><br/>
-        <b>Qwen Code</b><br/>
-        <sub>3+ models • Unlimited</sub>
-      </td>
-      <td align="center" width="150">
-        <img src="./public/providers/gemini-cli.png" width="70" alt="Gemini CLI"/><br/>
-        <b>Gemini CLI</b><br/>
-        <sub>180K/month FREE</sub>
-      </td>
      <td align="center" width="150">
        <img src="./public/providers/kiro.png" width="70" alt="Kiro"/><br/>
        <b>Kiro AI</b><br/>
-        <sub>Claude • Unlimited</sub>
+        <sub>Claude 4.5 + GLM-5 + MiniMax<br/>Unlimited FREE</sub>
+      </td>
+      <td align="center" width="150">
+        <img src="./public/providers/opencode.png" width="70" alt="OpenCode Free"/><br/>
+        <b>OpenCode Free</b><br/>
+        <sub>No auth • Auto-fetch models<br/>Unlimited FREE</sub>
+      </td>
+      <td align="center" width="150">
+        <img src="./public/providers/gemini.png" width="70" alt="Vertex AI"/><br/>
+        <b>Vertex AI</b><br/>
+        <sub>Gemini 3 Pro + GLM-5 + DeepSeek<br/>$300 credits free</sub>
      </td>
    </tr>
  </table>
 </div>

+> **Note:** iFlow, Qwen and Gemini CLI free tiers were discontinued in 2026. Use Kiro / OpenCode Free / Vertex instead.
+
 ### 🔑 API Key Providers (40+)

 <div align="center">
@@ -349,9 +349,10 @@ Default URLs:

 | Feature | What It Does | Why It Matters |
 |---------|--------------|----------------|
+| 🚀 **RTK Token Saver** | Auto-compress tool_result content (git-diff, grep, find, ls, tree...) before sending to LLM | Save 20-40% tokens per request, keep more context window |
 | 🎯 **Smart 3-Tier Fallback** | Auto-route: Subscription → Cheap → Free | Never stop coding, zero downtime |
 | 📊 **Real-Time Quota Tracking** | Live token count + reset countdown | Maximize subscription value |
-| 🔄 **Format Translation** | OpenAI ↔ Claude ↔ Gemini seamless | Works with any CLI tool |
+| 🔄 **Format Translation** | OpenAI ↔ Claude ↔ Gemini ↔ Cursor ↔ Kiro ↔ Vertex | Works with any CLI tool |
 | 👥 **Multi-Account Support** | Multiple accounts per provider | Load balancing + redundancy |
 | 🔄 **Auto Token Refresh** | OAuth tokens refresh automatically | No manual re-login needed |
 | 🎨 **Custom Combos** | Create unlimited model combinations | Tailor fallback to your needs |
@@ -363,6 +364,21 @@ Default URLs:
 <details>
 <summary><b>📖 Feature Details</b></summary>

+### 🚀 RTK Token Saver
+
+Tool outputs (`git diff`, `grep`, `find`, `ls`, `tree`, log dumps...) often eat 30-50% of your prompt budget. RTK detects them and applies smart, lossless compression **before** the request hits the LLM:
+
+- **Filters:** `git-diff`, `git-status`, `grep`, `find`, `ls`, `tree`, `dedup-log`, `smart-truncate`, `read-numbered`, `search-list`
+- **Auto-detect:** No config needed — RTK peeks the first 1KB of each `tool_result` and picks the right filter.
+- **Safe by design:** If a filter fails, throws, or makes output bigger, RTK silently keeps the original text. Errors never break your request.
+- **Universal:** Works across all formats (OpenAI, Claude, Gemini, Cursor, Kiro, OpenAI Responses) because it runs **before** any format translation.
+- **Default ON:** Toggle anytime in Dashboard → Endpoint settings.
+
+```
+Without RTK: 47K tokens sent to LLM
+With RTK:    28K tokens sent to LLM   (40% saved · same context · same answer)
+```
+
 ### 🎯 Smart 3-Tier Fallback

 Create combos with automatic fallback:
@@ -386,7 +402,7 @@ Combo: "my-coding-stack"
 ### 🔄 Format Translation

 Seamless translation between formats:
- **OpenAI** ↔ **Claude** ↔ **Gemini** ↔ **OpenAI Responses**
+- **OpenAI** ↔ **Claude** ↔ **Gemini** ↔ **Cursor** ↔ **Kiro** ↔ **Vertex** ↔ **Antigravity** ↔ **Ollama** ↔ **OpenAI Responses**
 - Your CLI tool sends OpenAI format → 9Router translates → Provider receives native format
 - Works with any tool that supports custom OpenAI endpoints

@@ -464,18 +480,19 @@ Seamless translation between formats:

 | Tier | Provider | Cost | Quota Reset | Best For |
 |------|----------|------|-------------|----------|
-| **💳 SUBSCRIPTION** | Claude Code (Pro) | $20/mo | 5h + weekly | Already subscribed |
+| **🚀 TOKEN SAVER** | **RTK (built-in)** | **FREE** | Always on | **Save 20-40% tokens on EVERY request** |
+| **💳 SUBSCRIPTION** | Claude Code (Pro/Max) | $20-200/mo | 5h + weekly | Already subscribed |
 | | Codex (Plus/Pro) | $20-200/mo | 5h + weekly | OpenAI users |
-| | Gemini CLI | **FREE** | 180K/mo + 1K/day | Everyone! |
 | | GitHub Copilot | $10-19/mo | Monthly | GitHub users |
-| **💰 CHEAP** | GLM-4.7 | $0.6/1M | Daily 10AM | Budget backup |
-| | MiniMax M2.1 | $0.2/1M | 5-hour rolling | Cheapest option |
-| | Kimi K2 | $9/mo flat | 10M tokens/mo | Predictable cost |
-| **🆓 FREE** | iFlow | $0 | Unlimited | 8 models free |
-| | Qwen | $0 | Unlimited | 3 models free |
-| | Kiro | $0 | Unlimited | Claude free |
+| | Cursor IDE | $20/mo | Monthly | Cursor users |
+| **💰 CHEAP** | GLM-5.1 / GLM-4.7 | $0.6/1M | Daily 10AM | Budget backup |
+| | MiniMax M2.7 | $0.2/1M | 5-hour rolling | Cheapest option |
+| | Kimi K2.5 | $9/mo flat | 10M tokens/mo | Predictable cost |
+| **🆓 FREE** | Kiro AI | $0 | Unlimited | Claude 4.5 + GLM-5 + MiniMax free |
+| | OpenCode Free | $0 | Unlimited | No auth, auto-fetch models |
+| | Vertex AI | $300 credits | New GCP accounts | Gemini 3 Pro + DeepSeek + GLM-5 |

-**💡 Pro Tip:** Start with Gemini CLI (180K free/month) + iFlow (unlimited free) combo = $0 cost!
+**💡 Pro Tip:** RTK + Kiro AI + OpenCode Free combo = **$0 cost + 20-40% token savings**!

 ---

@@ -523,9 +540,9 @@ Reality Check:
 **Solution:**
 ```
 Combo: "maximize-claude"
-  1. cc/claude-opus-4-6        (use subscription fully)
-  2. glm/glm-4.7               (cheap backup when quota out)
-  3. if/kimi-k2-thinking       (free emergency fallback)
+  1. cc/claude-opus-4-7        (use subscription fully)
+  2. glm/glm-5.1               (cheap backup when quota out)
+  3. kr/claude-sonnet-4.5      (free emergency fallback)

 Monthly cost: $20 (subscription) + ~$5 (backup) = $25 total
 vs. $20 + hitting limits = frustration
@@ -538,12 +555,12 @@ vs. $20 + hitting limits = frustration
 **Solution:**
 ```
 Combo: "free-forever"
-  1. gc/gemini-3-flash         (180K free/month)
-  2. if/kimi-k2-thinking       (unlimited free)
-  3. qw/qwen3-coder-plus       (unlimited free)
+  1. kr/claude-sonnet-4.5      (Claude 4.5 free unlimited)
+  2. kr/glm-5                  (GLM-5 free via Kiro)
+  3. oc/<auto>                 (OpenCode Free, no auth)

 Monthly cost: $0
-Quality: Production-ready models
+Quality: Production-ready models + RTK saves 20-40% tokens
 ```

 ### Case 3: "I need 24/7 coding, no interruptions"
@@ -553,11 +570,11 @@ Quality: Production-ready models
 **Solution:**
 ```
 Combo: "always-on"
-  1. cc/claude-opus-4-6        (best quality)
-  2. cx/gpt-5.2-codex          (second subscription)
-  3. glm/glm-4.7               (cheap, resets daily)
-  4. minimax/MiniMax-M2.1      (cheapest, 5h reset)
-  5. if/kimi-k2-thinking       (free unlimited)
+  1. cc/claude-opus-4-7        (best quality)
+  2. cx/gpt-5.5                (second subscription)
+  3. glm/glm-5.1               (cheap, resets daily)
+  4. minimax/MiniMax-M2.7      (cheapest, 5h reset)
+  5. kr/claude-sonnet-4.5      (free unlimited)

 Result: 5 layers of fallback = zero downtime
 Monthly cost: $20-200 (subscriptions) + $10-20 (backup)
@@ -570,9 +587,9 @@ Monthly cost: $20-200 (subscriptions) + $10-20 (backup)
 **Solution:**
 ```
 Combo: "openclaw-free"
-  1. if/glm-4.7                (unlimited free)
-  2. if/minimax-m2.1           (unlimited free)
-  3. if/kimi-k2-thinking       (unlimited free)
+  1. kr/claude-sonnet-4.5      (Claude 4.5 free)
+  2. kr/glm-5                  (GLM-5 free)
+  3. kr/MiniMax-M2.5           (MiniMax free)

 Monthly cost: $0
 Access via: WhatsApp, Telegram, Slack, Discord, iMessage, Signal...
@@ -614,16 +631,19 @@ The cost display is a "savings tracker" to help you understand your usage patter
 <details>
 <summary><b>🆓 Are FREE providers really unlimited?</b></summary>

-**Yes!** Providers marked as FREE (iFlow, Kiro, Qwen) are genuinely unlimited with **no hidden charges**. 
+**Yes!** The current FREE providers (Kiro, OpenCode Free, Vertex) are genuinely free with **no hidden charges**.

 These are free services offered by those respective companies:
- **iFlow**: Free unlimited access to 8+ models via OAuth
- **Kiro**: Free unlimited Claude models via AWS Builder ID  
- **Qwen**: Free unlimited access to Qwen models via device auth
+- **Kiro AI**: Free unlimited Claude 4.5 + GLM-5 + MiniMax via AWS Builder ID / Google / GitHub OAuth
+- **OpenCode Free**: No-auth passthrough proxy, models auto-fetched from `opencode.ai/zen/v1/models`
+- **Vertex AI**: $300 free credits for new Google Cloud accounts (90 days)

 9Router just routes your requests to them - there's no "catch" or future billing. They're truly free services, and 9Router makes them easy to use with fallback support.

-**Note:** Some subscription providers (Antigravity, GitHub Copilot) may have free preview periods that could become paid later, but this would be clearly announced by those providers, not 9Router.
+**Discontinued free tiers (no longer recommended):**
+- ❌ **iFlow**: Was free unlimited, now changed to paid (2026)
+- ❌ **Qwen Code**: Free OAuth tier discontinued by Alibaba on 2026-04-15
+- ❌ **Gemini CLI**: Still works, but using it with non-CLI tools (Claude, Codex, Cursor...) may result in account bans — only use if you stick to Gemini CLI itself

 </details>

@@ -689,8 +709,9 @@ Dashboard → Providers → Connect Claude Code
 → 5-hour + weekly quota tracking

 Models:
+  cc/claude-opus-4-7
  cc/claude-opus-4-6
-  cc/claude-sonnet-4-5-20250929
+  cc/claude-sonnet-4-6
  cc/claude-haiku-4-5-20251001
 ```

@@ -704,24 +725,12 @@ Dashboard → Providers → Connect Codex
 → 5-hour + weekly reset

 Models:
+  cx/gpt-5.5
+  cx/gpt-5.4
+  cx/gpt-5.3-codex
  cx/gpt-5.2-codex
-  cx/gpt-5.1-codex-max
 ```

-### Gemini CLI (FREE 180K/month!)
-
-```bash
-Dashboard → Providers → Connect Gemini CLI
-→ Google OAuth
-→ 180K completions/month + 1K/day
-
-Models:
-  gc/gemini-3-flash-preview
-  gc/gemini-2.5-pro
-```
-
-**Best Value:** Huge free tier! Use this before paid tiers.
-
 ### GitHub Copilot

 ```bash
@@ -730,9 +739,24 @@ Dashboard → Providers → Connect GitHub
 → Monthly reset (1st of month)

 Models:
-  gh/gpt-5
-  gh/claude-4.5-sonnet
-  gh/gemini-3-pro
+  gh/gpt-5.4
+  gh/claude-opus-4.7
+  gh/claude-sonnet-4.6
+  gh/gemini-3.1-pro-preview
+  gh/grok-code-fast-1
+```
+
+### Cursor IDE
+
+```bash
+Dashboard → Providers → Connect Cursor
+→ OAuth login
+→ Monthly subscription
+
+Models:
+  cu/claude-4.6-opus-max
+  cu/claude-4.5-sonnet-thinking
+  cu/gpt-5.3-codex
 ```

 </details>
@@ -740,7 +764,7 @@ Models:
 <details>
 <summary><b>💰 Cheap Providers (Backup)</b></summary>

-### GLM-4.7 (Daily reset, $0.6/1M)
+### GLM-5.1 / GLM-4.7 (Daily reset, $0.6/1M)

 1. Sign up: [Zhipu AI](https://open.bigmodel.cn/)
 2. Get API key from Coding Plan
@@ -748,74 +772,83 @@ Models:
   - Provider: `glm`
   - API Key: `your-key`

-**Use:** `glm/glm-4.7`
+**Use:** `glm/glm-5.1`, `glm/glm-5`, `glm/glm-4.7`

 **Pro Tip:** Coding Plan offers 3× quota at 1/7 cost! Reset daily 10:00 AM.

-### MiniMax M2.1 (5h reset, $0.20/1M)
+### MiniMax M2.7 (5h reset, $0.20/1M)

 1. Sign up: [MiniMax](https://www.minimax.io/)
 2. Get API key
 3. Dashboard → Add API Key

-**Use:** `minimax/MiniMax-M2.1`
+**Use:** `minimax/MiniMax-M2.7`, `minimax/MiniMax-M2.5`

 **Pro Tip:** Cheapest option for long context (1M tokens)!

-### Kimi K2 ($9/month flat)
+### Kimi K2.5 ($9/month flat)

 1. Subscribe: [Moonshot AI](https://platform.moonshot.ai/)
 2. Get API key
 3. Dashboard → Add API Key

-**Use:** `kimi/kimi-latest`
+**Use:** `kimi/kimi-k2.5`, `kimi/kimi-k2.5-thinking`

 **Pro Tip:** Fixed $9/month for 10M tokens = $0.90/1M effective cost!

 </details>

 <details>
-<summary><b>🆓 FREE Providers (Emergency Backup)</b></summary>
+<summary><b>🆓 FREE Providers (Recommended)</b></summary>

-### iFlow (8 FREE models)
-
-```bash
-Dashboard → Connect iFlow
-→ iFlow OAuth login
-→ Unlimited usage
-
-Models:
-  if/kimi-k2-thinking
-  if/qwen3-coder-plus
-  if/glm-4.7
-  if/minimax-m2
-  if/deepseek-r1
-```
-
-### Qwen (3 FREE models)
-
-```bash
-Dashboard → Connect Qwen
-→ Device code authorization
-→ Unlimited usage
-
-Models:
-  qw/qwen3-coder-plus
-  qw/qwen3-coder-flash
-```
-
-### Kiro (Claude FREE)
+### Kiro AI (Claude 4.5 + GLM-5 + MiniMax FREE)

 ```bash
 Dashboard → Connect Kiro
-→ AWS Builder ID, AWS IAM Identity Center, Google, GitHub
+→ AWS Builder ID, AWS IAM Identity Center, Google, or GitHub
 → Unlimited usage

 Models:
  kr/claude-sonnet-4.5
  kr/claude-haiku-4.5
+  kr/glm-5
+  kr/MiniMax-M2.5
+  kr/qwen3-coder-next
+  kr/deepseek-3.2
 ```

+**Pro Tip:** Best free option for Claude. No API key, no payment, fully unlimited.
+
+### OpenCode Free (No auth, auto-fetch models)
+
+```bash
+Dashboard → Connect OpenCode Free
+→ No login required (passthrough proxy)
+→ Models auto-fetched from opencode.ai/zen/v1/models
+```
+
+**Pro Tip:** Fastest setup. Just connect and start coding.
+
+### Vertex AI ($300 free credits for new GCP accounts)
+
+```bash
+Dashboard → Connect Vertex AI
+→ Upload Google Cloud Service Account JSON
+→ Enable Vertex AI API in your GCP project
+
+Models:
+  vertex/gemini-3.1-pro-preview
+  vertex/gemini-3-flash-preview
+  vertex/gemini-2.5-flash
+
+Vertex Partner (Anthropic / DeepSeek / GLM / Qwen via Vertex):
+  vertex-partner/glm-5-maas
+  vertex-partner/deepseek-v3.2-maas
+  vertex-partner/qwen3-next-80b-a3b-thinking-maas
+```
+
+**Pro Tip:** New Google Cloud accounts get $300 credits free for 90 days. Plenty for daily coding.
+
 </details>

 <details>
@@ -828,9 +861,9 @@ Dashboard → Combos → Create New

 Name: premium-coding
 Models:
-  1. cc/claude-opus-4-6 (Subscription primary)
-  2. glm/glm-4.7 (Cheap backup, $0.6/1M)
-  3. minimax/MiniMax-M2.1 (Cheapest fallback, $0.20/1M)
+  1. cc/claude-opus-4-7 (Subscription primary)
+  2. glm/glm-5.1 (Cheap backup, $0.6/1M)
+  3. minimax/MiniMax-M2.7 (Cheapest fallback, $0.20/1M)

 Use in CLI: premium-coding

@@ -846,11 +879,11 @@ Monthly cost example (100M tokens):
 ```
 Name: free-combo
 Models:
-  1. gc/gemini-3-flash-preview (180K free/month)
-  2. if/kimi-k2-thinking (unlimited)
-  3. qw/qwen3-coder-plus (unlimited)
+  1. kr/claude-sonnet-4.5 (Claude 4.5 free unlimited)
+  2. kr/glm-5 (GLM-5 free via Kiro)
+  3. vertex/gemini-3.1-pro-preview ($300 free credits)

-Cost: $0 forever!
+Cost: $0 forever (+ 20-40% token savings via RTK)!
 ```

 </details>
@@ -864,7 +897,7 @@ Cost: $0 forever!
 Settings → Models → Advanced:
  OpenAI API Base URL: http://localhost:20128/v1
  OpenAI API Key: [from 9router dashboard]
-  Model: cc/claude-opus-4-6
+  Model: cc/claude-opus-4-7
 ```

 Or use combo: `premium-coding`
@@ -904,7 +937,7 @@ Dashboard → CLI Tools → OpenClaw → Select Model → Apply
  "agents": {
    "defaults": {
      "model": {
-        "primary": "9router/if/glm-4.7"
+        "primary": "9router/kr/claude-sonnet-4.5"
      }
    }
  },
@@ -916,8 +949,8 @@ Dashboard → CLI Tools → OpenClaw → Select Model → Apply
        "api": "openai-completions",
        "models": [
          {
-            "id": "if/glm-4.7",
-            "name": "glm-4.7"
+            "id": "kr/claude-sonnet-4.5",
+            "name": "Claude Sonnet 4.5 (Kiro Free)"
          }
        ]
      }
@@ -934,7 +967,7 @@ Dashboard → CLI Tools → OpenClaw → Select Model → Apply
 Provider: OpenAI Compatible
 Base URL: http://localhost:20128/v1
 API Key: [from dashboard]
-Model: cc/claude-opus-4-6
+Model: cc/claude-opus-4-7
 ```

 </details>
@@ -1057,40 +1090,62 @@ Notes:
 <summary><b>View all available models</b></summary>

 **Claude Code (`cc/`)** - Pro/Max:
+- `cc/claude-opus-4-7`
 - `cc/claude-opus-4-6`
+- `cc/claude-sonnet-4-6`
 - `cc/claude-sonnet-4-5-20250929`
 - `cc/claude-haiku-4-5-20251001`

 **Codex (`cx/`)** - Plus/Pro:
+- `cx/gpt-5.5`
+- `cx/gpt-5.4`
+- `cx/gpt-5.3-codex`
 - `cx/gpt-5.2-codex`
 - `cx/gpt-5.1-codex-max`

-**Gemini CLI (`gc/`)** - FREE:
- `gc/gemini-3-flash-preview`
- `gc/gemini-2.5-pro`
-
 **GitHub Copilot (`gh/`)**:
- `gh/gpt-5`
- `gh/claude-4.5-sonnet`
+- `gh/gpt-5.4`
+- `gh/claude-opus-4.7`
+- `gh/claude-sonnet-4.6`
+- `gh/gemini-3.1-pro-preview`
+- `gh/grok-code-fast-1`
+
+**Cursor (`cu/`)** - Subscription:
+- `cu/claude-4.6-opus-max`
+- `cu/claude-4.5-sonnet-thinking`
+- `cu/gpt-5.3-codex`
+- `cu/kimi-k2.5`

 **GLM (`glm/`)** - $0.6/1M:
+- `glm/glm-5.1`
+- `glm/glm-5`
 - `glm/glm-4.7`

 **MiniMax (`minimax/`)** - $0.2/1M:
- `minimax/MiniMax-M2.1`
+- `minimax/MiniMax-M2.7`
+- `minimax/MiniMax-M2.5`

-**iFlow (`if/`)** - FREE:
- `if/kimi-k2-thinking`
- `if/qwen3-coder-plus`
- `if/deepseek-r1`
+**Kimi (`kimi/`)** - $9/mo flat:
+- `kimi/kimi-k2.5`
+- `kimi/kimi-k2.5-thinking`

-**Qwen (`qw/`)** - FREE:
- `qw/qwen3-coder-plus`
- `qw/qwen3-coder-flash`
-
-**Kiro (`kr/`)** - FREE:
+**Kiro (`kr/`)** - FREE unlimited:
 - `kr/claude-sonnet-4.5`
 - `kr/claude-haiku-4.5`
+- `kr/glm-5`
+- `kr/MiniMax-M2.5`
+- `kr/qwen3-coder-next`
+- `kr/deepseek-3.2`
+
+**OpenCode Free (`oc/`)** - FREE no-auth:
+- Auto-fetched from `opencode.ai/zen/v1/models`
+
+**Vertex AI (`vertex/`)** - $300 free credits:
+- `vertex/gemini-3.1-pro-preview`
+- `vertex/gemini-3-flash-preview`
+- `vertex/gemini-2.5-flash`
+- `vertex-partner/glm-5-maas`
+- `vertex-partner/deepseek-v3.2-maas`

 </details>

@@ -1104,16 +1159,17 @@ Notes:

 **Rate limiting**
 - Subscription quota out → Fallback to GLM/MiniMax
- Add combo: `cc/claude-opus-4-6 → glm/glm-4.7 → if/kimi-k2-thinking`
+- Add combo: `cc/claude-opus-4-7 → glm/glm-5.1 → kr/claude-sonnet-4.5`

 **OAuth token expired**
 - Auto-refreshed by 9Router
 - If issues persist: Dashboard → Provider → Reconnect

 **High costs**
+- Enable RTK in Dashboard → Endpoint settings (default ON, saves 20-40% tokens)
 - Check usage stats in Dashboard
 - Switch primary model to GLM/MiniMax
- Use free tier (Gemini CLI, iFlow) for non-critical tasks
+- Use free tier (Kiro, OpenCode Free, Vertex) for non-critical tasks

 **Dashboard opens on wrong port**
 - Set `PORT=20128` and `NEXT_PUBLIC_BASE_URL=http://localhost:20128`
--- a/open-sse/executors/antigravity.js
+++ b/open-sse/executors/antigravity.js
@@ -114,11 +114,11 @@ export class AntigravityExecutor extends BaseExecutor {
    };
  }

-  async refreshCredentials(credentials, log) {
+  async refreshCredentials(credentials, log, proxyOptions = null) {
    if (!credentials.refreshToken) return null;

    try {
-      const response = await fetch(OAUTH_ENDPOINTS.google.token, {
+      const response = await proxyAwareFetch(OAUTH_ENDPOINTS.google.token, {
        method: "POST",
        headers: { "Content-Type": "application/x-www-form-urlencoded", "Accept": "application/json" },
        body: new URLSearchParams({
@@ -127,7 +127,7 @@ export class AntigravityExecutor extends BaseExecutor {
          client_id: this.config.clientId,
          client_secret: this.config.clientSecret
        })
-      });
+      }, proxyOptions);

      if (!response.ok) return null;

--- a/open-sse/executors/base.js
+++ b/open-sse/executors/base.js
@@ -85,7 +85,7 @@ export class BaseExecutor {
  }

  // Override in subclass for provider-specific refresh
-  async refreshCredentials(credentials, log) {
+  async refreshCredentials(credentials, log, proxyOptions = null) {
    return null;
  }

--- a/open-sse/executors/default.js
+++ b/open-sse/executors/default.js
@@ -3,6 +3,7 @@ import { PROVIDERS } from "../config/providers.js";
 import { OAUTH_ENDPOINTS, buildKimiHeaders } from "../config/appConstants.js";
 import { buildClineHeaders } from "../../src/shared/utils/clineAuth.js";
 import { getCachedClaudeHeaders } from "../utils/claudeHeaderCache.js";
+import { proxyAwareFetch } from "../utils/proxyFetch.js";

 export class DefaultExecutor extends BaseExecutor {
  constructor(provider) {
@@ -154,19 +155,19 @@ export class DefaultExecutor extends BaseExecutor {
    return headers;
  }

-  async refreshCredentials(credentials, log) {
+  async refreshCredentials(credentials, log, proxyOptions = null) {
    if (!credentials.refreshToken) return null;

    const refreshers = {
-      claude: () => this.refreshWithJSON(OAUTH_ENDPOINTS.anthropic.token, { grant_type: "refresh_token", refresh_token: credentials.refreshToken, client_id: PROVIDERS.claude.clientId }),
-      codex: () => this.refreshWithForm(OAUTH_ENDPOINTS.openai.token, { grant_type: "refresh_token", refresh_token: credentials.refreshToken, client_id: PROVIDERS.codex.clientId, scope: "openid profile email offline_access" }),
-      qwen: () => this.refreshWithForm(OAUTH_ENDPOINTS.qwen.token, { grant_type: "refresh_token", refresh_token: credentials.refreshToken, client_id: PROVIDERS.qwen.clientId }),
-      iflow: () => this.refreshIflow(credentials.refreshToken),
-      gemini: () => this.refreshGoogle(credentials.refreshToken),
-      kiro: () => this.refreshKiro(credentials.refreshToken),
-      cline: () => this.refreshCline(credentials.refreshToken),
-      "kimi-coding": () => this.refreshKimiCoding(credentials.refreshToken),
-      kilocode: () => this.refreshKilocode(credentials.refreshToken)
+      claude: () => this.refreshWithJSON(OAUTH_ENDPOINTS.anthropic.token, { grant_type: "refresh_token", refresh_token: credentials.refreshToken, client_id: PROVIDERS.claude.clientId }, proxyOptions),
+      codex: () => this.refreshWithForm(OAUTH_ENDPOINTS.openai.token, { grant_type: "refresh_token", refresh_token: credentials.refreshToken, client_id: PROVIDERS.codex.clientId, scope: "openid profile email offline_access" }, proxyOptions),
+      qwen: () => this.refreshWithForm(OAUTH_ENDPOINTS.qwen.token, { grant_type: "refresh_token", refresh_token: credentials.refreshToken, client_id: PROVIDERS.qwen.clientId }, proxyOptions),
+      iflow: () => this.refreshIflow(credentials.refreshToken, proxyOptions),
+      gemini: () => this.refreshGoogle(credentials.refreshToken, proxyOptions),
+      kiro: () => this.refreshKiro(credentials.refreshToken, proxyOptions),
+      cline: () => this.refreshCline(credentials.refreshToken, proxyOptions),
+      "kimi-coding": () => this.refreshKimiCoding(credentials.refreshToken, proxyOptions),
+      kilocode: () => this.refreshKilocode(credentials.refreshToken, proxyOptions)
    };

    const refresher = refreshers[this.provider];
@@ -182,69 +183,69 @@ export class DefaultExecutor extends BaseExecutor {
    }
  }

-  async refreshWithJSON(url, body) {
-    const response = await fetch(url, {
+  async refreshWithJSON(url, body, proxyOptions = null) {
+    const response = await proxyAwareFetch(url, {
      method: "POST",
      headers: { "Content-Type": "application/json", "Accept": "application/json" },
      body: JSON.stringify(body)
-    });
+    }, proxyOptions);
    if (!response.ok) return null;
    const tokens = await response.json();
    return { accessToken: tokens.access_token, refreshToken: tokens.refresh_token || body.refresh_token, expiresIn: tokens.expires_in };
  }

-  async refreshWithForm(url, params) {
-    const response = await fetch(url, {
+  async refreshWithForm(url, params, proxyOptions = null) {
+    const response = await proxyAwareFetch(url, {
      method: "POST",
      headers: { "Content-Type": "application/x-www-form-urlencoded", "Accept": "application/json" },
      body: new URLSearchParams(params)
-    });
+    }, proxyOptions);
    if (!response.ok) return null;
    const tokens = await response.json();
    return { accessToken: tokens.access_token, refreshToken: tokens.refresh_token || params.refresh_token, expiresIn: tokens.expires_in };
  }

-  async refreshIflow(refreshToken) {
+  async refreshIflow(refreshToken, proxyOptions = null) {
    const basicAuth = btoa(`${PROVIDERS.iflow.clientId}:${PROVIDERS.iflow.clientSecret}`);
-    const response = await fetch(OAUTH_ENDPOINTS.iflow.token, {
+    const response = await proxyAwareFetch(OAUTH_ENDPOINTS.iflow.token, {
      method: "POST",
      headers: { "Content-Type": "application/x-www-form-urlencoded", "Accept": "application/json", "Authorization": `Basic ${basicAuth}` },
      body: new URLSearchParams({ grant_type: "refresh_token", refresh_token: refreshToken, client_id: PROVIDERS.iflow.clientId, client_secret: PROVIDERS.iflow.clientSecret })
-    });
+    }, proxyOptions);
    if (!response.ok) return null;
    const tokens = await response.json();
    return { accessToken: tokens.access_token, refreshToken: tokens.refresh_token || refreshToken, expiresIn: tokens.expires_in };
  }

-  async refreshGoogle(refreshToken) {
-    const response = await fetch(OAUTH_ENDPOINTS.google.token, {
+  async refreshGoogle(refreshToken, proxyOptions = null) {
+    const response = await proxyAwareFetch(OAUTH_ENDPOINTS.google.token, {
      method: "POST",
      headers: { "Content-Type": "application/x-www-form-urlencoded", "Accept": "application/json" },
      body: new URLSearchParams({ grant_type: "refresh_token", refresh_token: refreshToken, client_id: this.config.clientId, client_secret: this.config.clientSecret })
-    });
+    }, proxyOptions);
    if (!response.ok) return null;
    const tokens = await response.json();
    return { accessToken: tokens.access_token, refreshToken: tokens.refresh_token || refreshToken, expiresIn: tokens.expires_in };
  }

-  async refreshKiro(refreshToken) {
-    const response = await fetch(PROVIDERS.kiro.tokenUrl, {
+  async refreshKiro(refreshToken, proxyOptions = null) {
+    const response = await proxyAwareFetch(PROVIDERS.kiro.tokenUrl, {
      method: "POST",
      headers: { "Content-Type": "application/json", "Accept": "application/json", "User-Agent": "kiro-cli/1.0.0" },
      body: JSON.stringify({ refreshToken })
-    });
+    }, proxyOptions);
    if (!response.ok) return null;
    const tokens = await response.json();
    return { accessToken: tokens.accessToken, refreshToken: tokens.refreshToken || refreshToken, expiresIn: tokens.expiresIn };
  }

-  async refreshCline(refreshToken) {
+  async refreshCline(refreshToken, proxyOptions = null) {
    console.log('[DEBUG] Refreshing Cline token, refreshToken length:', refreshToken?.length);
-    const response = await fetch("https://api.cline.bot/api/v1/auth/refresh", {
+    const response = await proxyAwareFetch("https://api.cline.bot/api/v1/auth/refresh", {
      method: "POST",
      headers: { "Content-Type": "application/json", "Accept": "application/json" },
      body: JSON.stringify({ refreshToken, grantType: "refresh_token", clientType: "extension" })
-    });
+    }, proxyOptions);
    console.log('[DEBUG] Cline refresh response status:', response.status);
    if (!response.ok) {
      const errorText = await response.text();
@@ -260,9 +261,9 @@ export class DefaultExecutor extends BaseExecutor {
    return { accessToken: data?.accessToken, refreshToken: data?.refreshToken || refreshToken, expiresIn };
  }

-  async refreshKimiCoding(refreshToken) {
+  async refreshKimiCoding(refreshToken, proxyOptions = null) {
    const kimiHeaders = buildKimiHeaders();
-    const response = await fetch("https://auth.kimi.com/api/oauth/token", {
+    const response = await proxyAwareFetch("https://auth.kimi.com/api/oauth/token", {
      method: "POST",
      headers: {
        "Content-Type": "application/x-www-form-urlencoded",
@@ -270,13 +271,13 @@ export class DefaultExecutor extends BaseExecutor {
        ...kimiHeaders
      },
      body: new URLSearchParams({ grant_type: "refresh_token", refresh_token: refreshToken, client_id: "17e5f671-d194-4dfb-9706-5516cb48c098" })
-    });
+    }, proxyOptions);
    if (!response.ok) return null;
    const tokens = await response.json();
    return { accessToken: tokens.access_token, refreshToken: tokens.refresh_token || refreshToken, expiresIn: tokens.expires_in };
  }

-  async refreshKilocode(refreshToken) {
+  async refreshKilocode(refreshToken, proxyOptions = null) {
    // Kilocode uses device code flow, no refresh token support
    return null;
  }
--- a/open-sse/executors/github.js
+++ b/open-sse/executors/github.js
@@ -271,9 +271,9 @@ export class GithubExecutor extends BaseExecutor {
    };
  }

-  async refreshCopilotToken(githubAccessToken, log) {
+  async refreshCopilotToken(githubAccessToken, log, proxyOptions = null) {
    try {
-      const response = await fetch("https://api.github.com/copilot_internal/v2/token", {
+      const response = await proxyAwareFetch("https://api.github.com/copilot_internal/v2/token", {
        headers: {
          "Authorization": `token ${githubAccessToken}`,
          "User-Agent": GITHUB_COPILOT.USER_AGENT,
@@ -282,7 +282,7 @@ export class GithubExecutor extends BaseExecutor {
          "Accept": "application/json",
          "x-github-api-version": GITHUB_COPILOT.API_VERSION
        }
-      });
+      }, proxyOptions);
      if (!response.ok) {
        const errorText = await response.text();
        log?.error?.("TOKEN", `Copilot token refresh failed: ${response.status} ${errorText}`);
@@ -297,7 +297,7 @@ export class GithubExecutor extends BaseExecutor {
    }
  }

-  async refreshGitHubToken(refreshToken, log) {
+  async refreshGitHubToken(refreshToken, log, proxyOptions = null) {
    try {
      const params = {
        grant_type: "refresh_token",
@@ -308,11 +308,11 @@ export class GithubExecutor extends BaseExecutor {
        params.client_secret = this.config.clientSecret;
      }

-      const response = await fetch(OAUTH_ENDPOINTS.github.token, {
+      const response = await proxyAwareFetch(OAUTH_ENDPOINTS.github.token, {
        method: "POST",
        headers: { "Content-Type": "application/x-www-form-urlencoded", "Accept": "application/json" },
        body: new URLSearchParams(params)
-      });
+      }, proxyOptions);
      if (!response.ok) return null;
      const tokens = await response.json();
      log?.info?.("TOKEN", "GitHub token refreshed");
@@ -323,13 +323,13 @@ export class GithubExecutor extends BaseExecutor {
    }
  }

-  async refreshCredentials(credentials, log) {
-    let copilotResult = await this.refreshCopilotToken(credentials.accessToken, log);
+  async refreshCredentials(credentials, log, proxyOptions = null) {
+    let copilotResult = await this.refreshCopilotToken(credentials.accessToken, log, proxyOptions);

    if (!copilotResult && credentials.refreshToken) {
-      const githubTokens = await this.refreshGitHubToken(credentials.refreshToken, log);
+      const githubTokens = await this.refreshGitHubToken(credentials.refreshToken, log, proxyOptions);
      if (githubTokens?.accessToken) {
-        copilotResult = await this.refreshCopilotToken(githubTokens.accessToken, log);
+        copilotResult = await this.refreshCopilotToken(githubTokens.accessToken, log, proxyOptions);
        if (copilotResult) {
          return { ...githubTokens, copilotToken: copilotResult.token, copilotTokenExpiresAt: copilotResult.expiresAt };
        }
--- a/open-sse/executors/kiro.js
+++ b/open-sse/executors/kiro.js
@@ -378,7 +378,7 @@ export class KiroExecutor extends BaseExecutor {
    });
  }

-  async refreshCredentials(credentials, log) {
+  async refreshCredentials(credentials, log, proxyOptions = null) {
    if (!credentials.refreshToken) return null;

    try {
@@ -386,7 +386,8 @@ export class KiroExecutor extends BaseExecutor {
      const result = await refreshKiroToken(
        credentials.refreshToken,
        credentials.providerSpecificData,
-        log
+        log,
+        proxyOptions
      );

      return result;
--- a/open-sse/handlers/fetch/index.js
+++ b/open-sse/handlers/fetch/index.js
@@ -0,0 +1,237 @@
+// Web Fetch handler — dispatches to firecrawl, jina-reader, tavily, exa
+// Returns normalized shape across all providers
+
+const DEFAULT_TIMEOUT_MS = 15000;
+const DEFAULT_FORMAT = "markdown";
+
+/**
+ * @typedef {Object} FetchResult
+ * @property {boolean} success
+ * @property {number} [status]
+ * @property {string} [error]
+ * @property {Object} [data]
+ */
+
+/**
+ * Fetch with timeout abort.
+ * @param {string} url
+ * @param {RequestInit} init
+ * @param {number} timeoutMs
+ */
+// Strip non-ASCII chars from header values (HTTP headers must be ByteString).
+function sanitizeHeaders(headers) {
+  if (!headers) return headers;
+  const out = {};
+  for (const [k, v] of Object.entries(headers)) {
+    out[k] = typeof v === "string" ? v.replace(/[^\x00-\xFF]/g, "").trim() : v;
+  }
+  return out;
+}
+
+async function tryFetch(url, init, timeoutMs) {
+  const ctrl = new AbortController();
+  const timer = setTimeout(() => ctrl.abort(), timeoutMs);
+  try {
+    const res = await fetch(url, { ...init, headers: sanitizeHeaders(init.headers), signal: ctrl.signal });
+    return { ok: true, res };
+  } catch (err) {
+    const isAbort = err?.name === "AbortError";
+    return { ok: false, timeout: isAbort, error: err?.message || String(err) };
+  } finally {
+    clearTimeout(timer);
+  }
+}
+
+function truncate(text, max) {
+  if (!text || typeof text !== "string") return text || "";
+  if (!max || max <= 0) return text;
+  return text.length > max ? text.slice(0, max) : text;
+}
+
+function parseJinaTitle(text) {
+  const m = String(text || "").match(/^\s*#\s+(.+)$/m);
+  return m ? m[1].trim() : null;
+}
+
+function buildData({ provider, url, title, format, text, costUsd, responseMs, upstreamMs }) {
+  return {
+    provider,
+    url,
+    title: title || null,
+    content: { format, text: text || "", length: (text || "").length },
+    metadata: { author: null, published_at: null, language: null },
+    usage: { fetch_cost_usd: costUsd ?? null },
+    metrics: { response_time_ms: responseMs, upstream_latency_ms: upstreamMs }
+  };
+}
+
+async function readJsonOrText(res) {
+  const ct = res.headers.get("content-type") || "";
+  if (ct.includes("application/json")) {
+    try { return { json: await res.json() }; } catch { return { text: "" }; }
+  }
+  return { text: await res.text() };
+}
+
+/**
+ * Main handler.
+ * @param {Object} params
+ * @param {string} params.url
+ * @param {string} [params.format]
+ * @param {number} [params.maxCharacters]
+ * @param {string} params.provider
+ * @param {Object} [params.providerConfig]
+ * @param {Object} [params.credentials]
+ * @param {Function} [params.log]
+ * @returns {Promise<FetchResult>}
+ */
+export async function handleFetchCore({ url, format, maxCharacters, provider, providerConfig, credentials, log }) {
+  if (!url || typeof url !== "string") {
+    return { success: false, status: 400, error: "url is required" };
+  }
+  if (!provider) {
+    return { success: false, status: 400, error: "provider is required" };
+  }
+
+  const fmt = format || DEFAULT_FORMAT;
+  const timeoutMs = providerConfig?.timeoutMs || DEFAULT_TIMEOUT_MS;
+  const apiKey = credentials?.apiKey || credentials?.key || credentials?.token || "";
+  const costPerQuery = providerConfig?.costPerQuery ?? null;
+  const startedAt = Date.now();
+
+  try {
+    if (provider === "firecrawl") {
+      return await runFirecrawl({ url, fmt, timeoutMs, apiKey, maxCharacters, costPerQuery, startedAt });
+    }
+    if (provider === "jina-reader") {
+      return await runJina({ url, fmt, timeoutMs, apiKey, maxCharacters, costPerQuery, startedAt });
+    }
+    if (provider === "tavily") {
+      return await runTavily({ url, fmt, timeoutMs, apiKey, maxCharacters, costPerQuery, startedAt });
+    }
+    if (provider === "exa") {
+      return await runExa({ url, fmt, timeoutMs, apiKey, maxCharacters, costPerQuery, startedAt });
+    }
+    return { success: false, status: 400, error: `Unsupported provider: ${provider}` };
+  } catch (err) {
+    log?.("fetch handler error:", err?.message || err);
+    return { success: false, status: 502, error: err?.message || "Internal fetch error" };
+  }
+}
+
+async function runFirecrawl({ url, fmt, timeoutMs, apiKey, maxCharacters, costPerQuery, startedAt }) {
+  const upstreamStart = Date.now();
+  const r = await tryFetch("https://api.firecrawl.dev/v1/scrape", {
+    method: "POST",
+    headers: {
+      "content-type": "application/json",
+      ...(apiKey ? { authorization: `Bearer ${apiKey}` } : {})
+    },
+    body: JSON.stringify({ url, formats: [fmt] })
+  }, timeoutMs);
+
+  if (!r.ok) {
+    return { success: false, status: r.timeout ? 504 : 502, error: r.error };
+  }
+  const upstreamMs = Date.now() - upstreamStart;
+  const { json } = await readJsonOrText(r.res);
+  if (!r.res.ok) {
+    return { success: false, status: r.res.status, error: json?.error || `Firecrawl error: ${r.res.status}` };
+  }
+  const d = json?.data || {};
+  const text = truncate(d.markdown || d.html || d.text || "", maxCharacters);
+  const title = d.metadata?.title || null;
+  return {
+    success: true,
+    data: buildData({
+      provider: "firecrawl", url, title, format: fmt, text,
+      costUsd: costPerQuery, responseMs: Date.now() - startedAt, upstreamMs
+    })
+  };
+}
+
+async function runJina({ url, fmt, timeoutMs, apiKey, maxCharacters, costPerQuery, startedAt }) {
+  const target = `https://r.jina.ai/${encodeURIComponent(url)}`;
+  const upstreamStart = Date.now();
+  const r = await tryFetch(target, {
+    method: "GET",
+    headers: apiKey ? { authorization: `Bearer ${apiKey}` } : {}
+  }, timeoutMs);
+
+  if (!r.ok) {
+    return { success: false, status: r.timeout ? 504 : 502, error: r.error };
+  }
+  const upstreamMs = Date.now() - upstreamStart;
+  const body = await r.res.text();
+  if (!r.res.ok) {
+    return { success: false, status: r.res.status, error: body?.slice(0, 500) || `Jina error: ${r.res.status}` };
+  }
+  const text = truncate(body, maxCharacters);
+  return {
+    success: true,
+    data: buildData({
+      provider: "jina-reader", url, title: parseJinaTitle(body), format: fmt, text,
+      costUsd: costPerQuery, responseMs: Date.now() - startedAt, upstreamMs
+    })
+  };
+}
+
+async function runTavily({ url, fmt, timeoutMs, apiKey, maxCharacters, costPerQuery, startedAt }) {
+  const upstreamStart = Date.now();
+  const r = await tryFetch("https://api.tavily.com/extract", {
+    method: "POST",
+    headers: {
+      "content-type": "application/json",
+      ...(apiKey ? { authorization: `Bearer ${apiKey}` } : {})
+    },
+    body: JSON.stringify({ urls: [url], extract_depth: "basic" })
+  }, timeoutMs);
+
+  if (!r.ok) {
+    return { success: false, status: r.timeout ? 504 : 502, error: r.error };
+  }
+  const upstreamMs = Date.now() - upstreamStart;
+  const { json } = await readJsonOrText(r.res);
+  if (!r.res.ok) {
+    return { success: false, status: r.res.status, error: json?.error || `Tavily error: ${r.res.status}` };
+  }
+  const first = json?.results?.[0] || {};
+  const text = truncate(first.raw_content || "", maxCharacters);
+  return {
+    success: true,
+    data: buildData({
+      provider: "tavily", url, title: null, format: fmt, text,
+      costUsd: costPerQuery, responseMs: Date.now() - startedAt, upstreamMs
+    })
+  };
+}
+
+async function runExa({ url, fmt, timeoutMs, apiKey, maxCharacters, costPerQuery, startedAt }) {
+  const upstreamStart = Date.now();
+  const r = await tryFetch("https://api.exa.ai/contents", {
+    method: "POST",
+    headers: {
+      "content-type": "application/json",
+      ...(apiKey ? { "x-api-key": apiKey } : {})
+    },
+    body: JSON.stringify({ ids: [url], text: true })
+  }, timeoutMs);
+
+  if (!r.ok) {
+    return { success: false, status: r.timeout ? 504 : 502, error: r.error };
+  }
+  const upstreamMs = Date.now() - upstreamStart;
+  const { json } = await readJsonOrText(r.res);
+  if (!r.res.ok) {
+    return { success: false, status: r.res.status, error: json?.error || `Exa error: ${r.res.status}` };
+  }
+  const first = json?.results?.[0] || {};
+  const text = truncate(first.text || "", maxCharacters);
+  return {
+    success: true,
+    data: buildData({
+      provider: "exa", url, title: first.title || null, format: fmt, text,
+      costUsd: costPerQuery, responseMs: Date.now() - startedAt, upstreamMs
+    })
+  };
+}
--- a/open-sse/handlers/search/callers.js
+++ b/open-sse/handlers/search/callers.js
@@ -0,0 +1,371 @@
+/**
+ * Search Provider Request Builders
+ *
+ * Ported from OmniRoute open-sse/handlers/search.ts (lines 223-610).
+ * Builds HTTP request `{ url, init }` for 10 search providers.
+ *
+ * @typedef {Object} SearchProviderConfig
+ * @property {string} id
+ * @property {string} baseUrl
+ * @property {string} [method]
+ *
+ * @typedef {Object} ContentOptions
+ * @property {boolean} [snippet]
+ * @property {boolean} [full_page]
+ * @property {string}  [format]
+ * @property {number}  [max_characters]
+ *
+ * @typedef {Object} SearchRequestParams
+ * @property {string}   query
+ * @property {string}   searchType
+ * @property {number}   maxResults
+ * @property {string}   [token]
+ * @property {string}   [country]
+ * @property {string}   [language]
+ * @property {string}   [timeRange]
+ * @property {number}   [offset]
+ * @property {string[]} [domainFilter]
+ * @property {ContentOptions}        [contentOptions]
+ * @property {Record<string,unknown>} [providerOptions]
+ * @property {Record<string,unknown>} [providerSpecificData]
+ */
+
+// ── Helpers ─────────────────────────────────────────────────────────────
+
+/**
+ * Split domain filter into includes / excludes (excludes prefixed with "-").
+ * @param {string[]} [domainFilter]
+ * @returns {{includes: string[], excludes: string[]}}
+ */
+export function parseDomainFilter(domainFilter) {
+  if (!domainFilter?.length) return { includes: [], excludes: [] };
+  const includes = domainFilter.filter((d) => !d.startsWith("-"));
+  const excludes = domainFilter.filter((d) => d.startsWith("-")).map((d) => d.slice(1));
+  return { includes, excludes };
+}
+
+/**
+ * Read string setting from providerOptions first, then providerSpecificData.
+ * @param {SearchRequestParams} params
+ * @param {string} key
+ * @returns {string|undefined}
+ */
+export function getProviderSetting(params, key) {
+  const fromOptions = params.providerOptions?.[key];
+  if (typeof fromOptions === "string" && fromOptions.trim().length > 0) {
+    return fromOptions.trim();
+  }
+  const fromProviderData = params.providerSpecificData?.[key];
+  if (typeof fromProviderData === "string" && fromProviderData.trim().length > 0) {
+    return fromProviderData.trim();
+  }
+  return undefined;
+}
+
+/**
+ * Resolve base URL with optional override from providerOptions.baseUrl.
+ * @param {SearchProviderConfig} config
+ * @param {SearchRequestParams} params
+ * @returns {string}
+ */
+export function resolveBaseUrl(config, params) {
+  const override = getProviderSetting(params, "baseUrl");
+  return (override || config.baseUrl).replace(/\/+$/, "");
+}
+
+/**
+ * Convert offset+maxResults to 1-indexed page number.
+ * @param {number|undefined} offset
+ * @param {number} maxResults
+ * @returns {number|undefined}
+ */
+export function toPageNumber(offset, maxResults) {
+  if (typeof offset !== "number" || offset <= 0 || maxResults <= 0) return undefined;
+  return Math.floor(offset / maxResults) + 1;
+}
+
+// ── Provider Request Builders ───────────────────────────────────────────
+
+function buildSerperRequest(config, params) {
+  const endpoint = params.searchType === "news" ? "/news" : "/search";
+  const body = { q: params.query, num: params.maxResults };
+  if (params.country) body.gl = params.country.toLowerCase();
+  if (params.language) body.hl = params.language;
+  return {
+    url: `${resolveBaseUrl(config, params)}${endpoint}`,
+    init: {
+      method: "POST",
+      headers: { "Content-Type": "application/json", "X-API-Key": params.token },
+      body: JSON.stringify(body),
+    },
+  };
+}
+
+function buildBraveRequest(config, params) {
+  const endpoint = params.searchType === "news" ? "/news/search" : "/web/search";
+  const qp = new URLSearchParams({ q: params.query, count: String(params.maxResults) });
+  if (params.country) qp.set("country", params.country);
+  if (params.language) qp.set("search_lang", params.language);
+  return {
+    url: `${resolveBaseUrl(config, params)}${endpoint}?${qp}`,
+    init: {
+      method: "GET",
+      headers: { Accept: "application/json", "X-Subscription-Token": params.token },
+    },
+  };
+}
+
+function buildPerplexityRequest(config, params) {
+  const body = { query: params.query, max_results: params.maxResults };
+  if (params.country) body.country = params.country;
+  if (params.language) body.search_language_filter = [params.language];
+  if (params.domainFilter?.length) body.search_domain_filter = params.domainFilter;
+  return {
+    url: resolveBaseUrl(config, params),
+    init: {
+      method: "POST",
+      headers: { "Content-Type": "application/json", Authorization: `Bearer ${params.token}` },
+      body: JSON.stringify(body),
+    },
+  };
+}
+
+function buildExaRequest(config, params) {
+  const { includes, excludes } = parseDomainFilter(params.domainFilter);
+  const body = {
+    query: params.query,
+    numResults: params.maxResults,
+    type: "auto",
+    text: true,
+    highlights: true,
+  };
+  if (includes.length) body.includeDomains = includes;
+  if (excludes.length) body.excludeDomains = excludes;
+  if (params.searchType === "news") body.category = "news";
+  return {
+    url: resolveBaseUrl(config, params),
+    init: {
+      method: "POST",
+      headers: { "Content-Type": "application/json", "x-api-key": params.token },
+      body: JSON.stringify(body),
+    },
+  };
+}
+
+function buildTavilyRequest(config, params) {
+  const { includes, excludes } = parseDomainFilter(params.domainFilter);
+  const body = {
+    query: params.query,
+    max_results: params.maxResults,
+    topic: params.searchType === "news" ? "news" : "general",
+  };
+  if (includes.length) body.include_domains = includes;
+  if (excludes.length) body.exclude_domains = excludes;
+  if (params.country) body.country = params.country;
+  return {
+    url: resolveBaseUrl(config, params),
+    init: {
+      method: "POST",
+      headers: { "Content-Type": "application/json", Authorization: `Bearer ${params.token}` },
+      body: JSON.stringify(body),
+    },
+  };
+}
+
+function buildGooglePseRequest(config, params) {
+  const apiKey = params.token;
+  const cx = getProviderSetting(params, "cx");
+  if (!apiKey || !cx) {
+    throw new Error("Google Programmable Search requires both apiKey and cx");
+  }
+  const qp = new URLSearchParams({
+    key: apiKey,
+    cx,
+    q: params.query,
+    num: String(Math.min(params.maxResults, 10)),
+  });
+  if (params.country) qp.set("gl", params.country.toLowerCase());
+  if (params.language) qp.set("hl", params.language);
+  if (params.timeRange && params.timeRange !== "any") {
+    const dateRestrictMap = { day: "d1", week: "w1", month: "m1", year: "y1" };
+    const dateRestrict = dateRestrictMap[params.timeRange];
+    if (dateRestrict) qp.set("dateRestrict", dateRestrict);
+  }
+  if (typeof params.offset === "number" && params.offset > 0) {
+    qp.set("start", String(Math.min(params.offset + 1, 91)));
+  }
+  return {
+    url: `${resolveBaseUrl(config, params)}?${qp}`,
+    init: {
+      method: "GET",
+      headers: { Accept: "application/json" },
+    },
+  };
+}
+
+function buildLinkupRequest(config, params) {
+  const apiKey = params.token;
+  if (!apiKey) throw new Error("Linkup Search requires an API key");
+
+  const { includes, excludes } = parseDomainFilter(params.domainFilter);
+  const requestedDepth = getProviderSetting(params, "depth");
+  const depth =
+    requestedDepth && ["fast", "standard", "deep"].includes(requestedDepth)
+      ? requestedDepth
+      : "standard";
+
+  const body = {
+    q: params.query,
+    depth,
+    outputType: "searchResults",
+    maxResults: params.maxResults,
+  };
+  if (includes.length) body.includeDomains = includes;
+  if (excludes.length) body.excludeDomains = excludes;
+  if (params.timeRange && params.timeRange !== "any") {
+    const today = new Date();
+    const toDate = today.toISOString().slice(0, 10);
+    const from = new Date(today);
+    if (params.timeRange === "day") from.setUTCDate(from.getUTCDate() - 1);
+    if (params.timeRange === "week") from.setUTCDate(from.getUTCDate() - 7);
+    if (params.timeRange === "month") from.setUTCMonth(from.getUTCMonth() - 1);
+    if (params.timeRange === "year") from.setUTCFullYear(from.getUTCFullYear() - 1);
+    body.fromDate = from.toISOString().slice(0, 10);
+    body.toDate = toDate;
+  }
+
+  return {
+    url: resolveBaseUrl(config, params),
+    init: {
+      method: "POST",
+      headers: { "Content-Type": "application/json", Authorization: `Bearer ${apiKey}` },
+      body: JSON.stringify(body),
+    },
+  };
+}
+
+function buildSearchApiRequest(config, params) {
+  const apiKey = params.token;
+  if (!apiKey) throw new Error("SearchAPI requires an API key");
+
+  const qp = new URLSearchParams({
+    engine: params.searchType === "news" ? "google_news" : "google",
+    q: params.query,
+    api_key: apiKey,
+  });
+  if (params.country) qp.set("gl", params.country.toLowerCase());
+  if (params.language) qp.set("hl", params.language);
+
+  const page = toPageNumber(params.offset, params.maxResults);
+  if (page) qp.set("page", String(page));
+
+  return {
+    url: `${resolveBaseUrl(config, params)}?${qp}`,
+    init: {
+      method: "GET",
+      headers: { Accept: "application/json" },
+    },
+  };
+}
+
+function buildYouComRequest(config, params) {
+  const apiKey = params.token;
+  if (!apiKey) throw new Error("You.com Search requires an API key");
+
+  const { includes, excludes } = parseDomainFilter(params.domainFilter);
+  const qp = new URLSearchParams({
+    query: params.query,
+    count: String(Math.min(params.maxResults, 100)),
+  });
+
+  if (params.timeRange && params.timeRange !== "any") qp.set("freshness", params.timeRange);
+  if (typeof params.offset === "number" && params.offset > 0 && params.maxResults > 0) {
+    qp.set("offset", String(Math.min(Math.floor(params.offset / params.maxResults), 9)));
+  }
+  if (params.country) qp.set("country", params.country);
+  if (params.language) qp.set("language", params.language);
+  if (includes.length) qp.set("include_domains", includes.join(","));
+  if (excludes.length) qp.set("exclude_domains", excludes.join(","));
+
+  if (params.contentOptions?.full_page) {
+    qp.set("livecrawl", params.searchType === "news" ? "news" : "web");
+    qp.append(
+      "livecrawl_formats",
+      params.contentOptions.format === "markdown" ? "markdown" : "html"
+    );
+  }
+
+  return {
+    url: `${resolveBaseUrl(config, params)}?${qp}`,
+    init: {
+      method: "GET",
+      headers: { Accept: "application/json", "X-API-Key": apiKey },
+    },
+  };
+}
+
+function buildSearxngRequest(config, params) {
+  const baseUrl = resolveBaseUrl(config, params);
+  const url = baseUrl.endsWith("/search") ? baseUrl : `${baseUrl}/search`;
+  const qp = new URLSearchParams({
+    q: params.query,
+    format: "json",
+    categories: params.searchType === "news" ? "news" : "general",
+  });
+  if (params.language) qp.set("language", params.language);
+  if (params.timeRange && params.timeRange !== "any") qp.set("time_range", params.timeRange);
+
+  const page = toPageNumber(params.offset, params.maxResults);
+  if (page) qp.set("pageno", String(page));
+
+  return {
+    url: `${url}?${qp}`,
+    init: {
+      method: "GET",
+      headers: { Accept: "application/json" },
+    },
+  };
+}
+
+// ── Dispatcher ──────────────────────────────────────────────────────────
+
+const BUILDERS = {
+  "serper": buildSerperRequest,
+  "brave-search": buildBraveRequest,
+  "perplexity": buildPerplexityRequest,
+  "exa": buildExaRequest,
+  "tavily": buildTavilyRequest,
+  "google-pse": buildGooglePseRequest,
+  "linkup": buildLinkupRequest,
+  "searchapi": buildSearchApiRequest,
+  "youcom": buildYouComRequest,
+  "searxng": buildSearxngRequest,
+};
+
+/**
+ * Dispatch to the correct provider builder by `provider.id`.
+ * Falls back to generic POST + bearer auth for unknown providers.
+ * @param {SearchProviderConfig} provider
+ * @param {SearchRequestParams} params
+ * @returns {{url: string, init: RequestInit}}
+ */
+export function buildSearchRequest(provider, params) {
+  const builder = BUILDERS[provider.id];
+  if (builder) return builder(provider, params);
+
+  return {
+    url: resolveBaseUrl(provider, params),
+    init: {
+      method: provider.method || "POST",
+      headers: {
+        "Content-Type": "application/json",
+        ...(params.token ? { Authorization: `Bearer ${params.token}` } : {}),
+      },
+      body: JSON.stringify({
+        query: params.query,
+        max_results: params.maxResults,
+        search_type: params.searchType,
+      }),
+    },
+  };
+}
--- a/open-sse/handlers/search/chatSearch.js
+++ b/open-sse/handlers/search/chatSearch.js
@@ -0,0 +1,409 @@
+/**
+ * Wrap chat-completions endpoints (with built-in web search) into the unified
+ * /v1/search response format. Supports gemini, openai, xai, kimi, minimax, perplexity.
+ */
+
+const REQUEST_TIMEOUT_MS = 15000;
+const DEFAULT_MAX_RESULTS = 10;
+
+/**
+ * Normalize a citation entry into the unified result shape.
+ * @param {{url:string, title?:string, snippet?:string}} c
+ * @param {number} index
+ * @param {string} provider
+ * @param {string} retrievedAt
+ */
+function toResult(c, index, provider, retrievedAt) {
+  return {
+    title: c.title || "",
+    url: c.url,
+    snippet: c.snippet || "",
+    position: index + 1,
+    score: null,
+    published_at: null,
+    favicon_url: null,
+    content: null,
+    metadata: {},
+    citation: { provider, retrieved_at: retrievedAt, rank: index + 1 },
+    provider_raw: null
+  };
+}
+
+/** Coerce a citation that might be a raw URL string or an object. */
+function normalizeCitation(c) {
+  if (!c) return null;
+  if (typeof c === "string") return { url: c };
+  if (typeof c === "object" && c.url) return c;
+  return null;
+}
+
+/**
+ * Provider-specific configuration map. All providers must implement:
+ * { endpoint, defaultModel, buildBody, buildHeaders, extractAnswer }
+ */
+const CHAT_SEARCH_CONFIG = {
+  gemini: {
+    endpoint: (model) =>
+      `https://generativelanguage.googleapis.com/v1beta/models/${model}:generateContent`,
+    defaultModel: "gemini-2.5-flash",
+    buildBody: (query) => ({
+      contents: [{ role: "user", parts: [{ text: query }] }],
+      tools: [{ google_search: {} }]
+    }),
+    buildHeaders: (token) => ({
+      "Content-Type": "application/json",
+      "x-goog-api-key": token
+    }),
+    extractAnswer: (data) => {
+      const candidate = data?.candidates?.[0];
+      const parts = candidate?.content?.parts || [];
+      const text = parts.map((p) => p?.text || "").filter(Boolean).join("");
+      const chunks = candidate?.groundingMetadata?.groundingChunks || [];
+      const citations = chunks
+        .map((ch) => ch?.web)
+        .filter(Boolean)
+        .map((w) => ({ url: w.uri || w.url, title: w.title || "" }))
+        .filter((c) => c.url);
+      const tokens = data?.usageMetadata?.totalTokenCount || 0;
+      return { text, citations, tokens };
+    }
+  },
+
+  openai: {
+    endpoint: () => "https://api.openai.com/v1/chat/completions",
+    defaultModel: "gpt-4o-mini",
+    buildBody: (query, model) => {
+      const body = {
+        model,
+        messages: [{ role: "user", content: query }]
+      };
+      // Non-search-preview models need explicit web_search tool
+      if (!/search/i.test(model)) {
+        body.tools = [{ type: "web_search" }];
+      }
+      return body;
+    },
+    buildHeaders: (token) => ({
+      "Content-Type": "application/json",
+      Authorization: `Bearer ${token}`
+    }),
+    extractAnswer: (data) => {
+      const msg = data?.choices?.[0]?.message || {};
+      const text = msg.content || "";
+      const annotations = Array.isArray(msg.annotations) ? msg.annotations : [];
+      const fromAnn = annotations
+        .map((a) => a?.url_citation)
+        .filter(Boolean)
+        .map((u) => ({ url: u.url, title: u.title || "" }));
+      const fromTop = Array.isArray(data?.citations)
+        ? data.citations.map(normalizeCitation).filter(Boolean)
+        : [];
+      const citations = fromAnn.length ? fromAnn : fromTop;
+      const tokens = data?.usage?.total_tokens || 0;
+      return { text, citations, tokens };
+    }
+  },
+
+  xai: {
+    endpoint: () => "https://api.x.ai/v1/responses",
+    defaultModel: "grok-4.20-reasoning",
+    buildBody: (query, model) => ({
+      model,
+      input: [{ role: "user", content: query }],
+      tools: [{ type: "web_search" }]
+    }),
+    buildHeaders: (token) => ({
+      "Content-Type": "application/json",
+      Authorization: `Bearer ${token}`
+    }),
+    extractAnswer: (data) => {
+      // /v1/responses returns output[] array of message/tool blocks
+      const output = Array.isArray(data?.output) ? data.output : [];
+      let text = "";
+      const citations = [];
+      for (const item of output) {
+        const parts = Array.isArray(item?.content) ? item.content : [];
+        for (const p of parts) {
+          if (typeof p?.text === "string") text += p.text;
+          const anns = Array.isArray(p?.annotations) ? p.annotations : [];
+          for (const a of anns) {
+            const c = normalizeCitation(a?.url ? a : a?.url_citation);
+            if (c) citations.push(c);
+          }
+        }
+      }
+      // Fallback: top-level citations array (some response variants)
+      if (!citations.length && Array.isArray(data?.citations)) {
+        for (const c of data.citations) {
+          const n = normalizeCitation(c);
+          if (n) citations.push(n);
+        }
+      }
+      const tokens = data?.usage?.total_tokens || 0;
+      return { text, citations, tokens };
+    }
+  },
+
+  kimi: {
+    endpoint: () => "https://api.moonshot.cn/v1/chat/completions",
+    defaultModel: "kimi-k2.5",
+    buildBody: (query, model) => ({
+      model,
+      messages: [{ role: "user", content: query }],
+      tools: [
+        { type: "builtin_function", function: { name: "$web_search" } }
+      ]
+    }),
+    buildHeaders: (token) => ({
+      "Content-Type": "application/json",
+      Authorization: `Bearer ${token}`
+    }),
+    extractAnswer: (data) => {
+      const msg = data?.choices?.[0]?.message || {};
+      const text = msg.content || "";
+      const calls = Array.isArray(msg.tool_calls) ? msg.tool_calls : [];
+      const citations = [];
+      for (const call of calls) {
+        const argStr = call?.function?.arguments;
+        if (!argStr) continue;
+        let parsed;
+        try {
+          parsed = typeof argStr === "string" ? JSON.parse(argStr) : argStr;
+        } catch {
+          continue;
+        }
+        const items =
+          parsed?.search_results ||
+          parsed?.results ||
+          parsed?.references ||
+          [];
+        if (Array.isArray(items)) {
+          for (const it of items) {
+            const url = it?.url || it?.link;
+            if (!url) continue;
+            citations.push({
+              url,
+              title: it.title || "",
+              snippet: it.snippet || it.summary || ""
+            });
+          }
+        }
+      }
+      const tokens = data?.usage?.total_tokens || 0;
+      return { text, citations, tokens };
+    }
+  },
+
+  minimax: {
+    endpoint: () => "https://api.minimaxi.com/v1/text/chatcompletion_v2",
+    defaultModel: "MiniMax-M2.7",
+    buildBody: (query, model) => ({
+      model,
+      messages: [{ role: "user", content: query }],
+      tools: [{ type: "web_search" }]
+    }),
+    buildHeaders: (token) => ({
+      "Content-Type": "application/json",
+      Authorization: `Bearer ${token}`
+    }),
+    extractAnswer: (data) => {
+      const msg = data?.choices?.[0]?.message || {};
+      const text = msg.content || "";
+      const citations = [];
+      const direct = Array.isArray(data?.web_search_results)
+        ? data.web_search_results
+        : [];
+      for (const it of direct) {
+        const url = it?.url || it?.link;
+        if (url) {
+          citations.push({
+            url,
+            title: it.title || "",
+            snippet: it.snippet || it.summary || ""
+          });
+        }
+      }
+      if (!citations.length) {
+        const calls = Array.isArray(msg.tool_calls) ? msg.tool_calls : [];
+        for (const call of calls) {
+          const argStr = call?.function?.arguments;
+          if (!argStr) continue;
+          let parsed;
+          try {
+            parsed = typeof argStr === "string" ? JSON.parse(argStr) : argStr;
+          } catch {
+            continue;
+          }
+          const items = parsed?.results || parsed?.search_results || [];
+          if (Array.isArray(items)) {
+            for (const it of items) {
+              const url = it?.url || it?.link;
+              if (!url) continue;
+              citations.push({
+                url,
+                title: it.title || "",
+                snippet: it.snippet || ""
+              });
+            }
+          }
+        }
+      }
+      const tokens = data?.usage?.total_tokens || 0;
+      return { text, citations, tokens };
+    }
+  },
+
+  perplexity: {
+    endpoint: () => "https://api.perplexity.ai/chat/completions",
+    defaultModel: "sonar",
+    buildBody: (query, model) => ({
+      model,
+      messages: [{ role: "user", content: query }]
+    }),
+    buildHeaders: (token) => ({
+      "Content-Type": "application/json",
+      Authorization: `Bearer ${token}`
+    }),
+    extractAnswer: (data) => {
+      const msg = data?.choices?.[0]?.message || {};
+      const text = msg.content || "";
+      const raw = data?.citations || [];
+      const citations = Array.isArray(raw)
+        ? raw.map(normalizeCitation).filter(Boolean)
+        : [];
+      const tokens = data?.usage?.total_tokens || 0;
+      return { text, citations, tokens };
+    }
+  }
+};
+
+/**
+ * Execute a chat-search request against the chosen provider.
+ * @param {object} params
+ * @param {string} params.provider
+ * @param {string} params.query
+ * @param {number} [params.maxResults]
+ * @param {string} [params.model]
+ * @param {{apiKey?:string, accessToken?:string}} params.credentials
+ * @param {{info?:Function, warn?:Function, error?:Function}} [params.log]
+ * @returns {Promise<{success:boolean, status?:number, error?:string, data?:object}>}
+ */
+export async function handleChatSearch({
+  provider,
+  query,
+  maxResults,
+  model,
+  credentials,
+  log
+}) {
+  const startTime = Date.now();
+  const cfg = CHAT_SEARCH_CONFIG[provider];
+
+  if (!cfg) {
+    return {
+      success: false,
+      status: 400,
+      error: `Unsupported chat-search provider: ${provider}`
+    };
+  }
+
+  if (!query || typeof query !== "string") {
+    return { success: false, status: 400, error: "Missing query" };
+  }
+
+  const token = credentials?.apiKey || credentials?.accessToken;
+  if (!token) {
+    return {
+      success: false,
+      status: 401,
+      error: "Missing credentials (apiKey or accessToken)"
+    };
+  }
+
+  const limit =
+    Number.isFinite(maxResults) && maxResults > 0
+      ? Math.floor(maxResults)
+      : DEFAULT_MAX_RESULTS;
+  const useModel = model || cfg.defaultModel;
+  const url = cfg.endpoint(useModel);
+  const body = cfg.buildBody(query, useModel);
+  const headers = cfg.buildHeaders(token);
+
+  const controller = new AbortController();
+  const timer = setTimeout(() => controller.abort(), REQUEST_TIMEOUT_MS);
+
+  let upstreamStart = Date.now();
+  let resp;
+  try {
+    resp = await fetch(url, {
+      method: "POST",
+      headers,
+      body: JSON.stringify(body),
+      signal: controller.signal
+    });
+  } catch (err) {
+    clearTimeout(timer);
+    if (err?.name === "AbortError") {
+      log?.warn?.(`[chatSearch] timeout provider=${provider}`);
+      return { success: false, status: 504, error: "Upstream timeout" };
+    }
+    log?.error?.(`[chatSearch] network error provider=${provider}: ${err?.message}`);
+    return {
+      success: false,
+      status: 502,
+      error: `Network error: ${err?.message || "unknown"}`
+    };
+  }
+  clearTimeout(timer);
+  const upstreamLatency = Date.now() - upstreamStart;
+
+  let data;
+  try {
+    data = await resp.json();
+  } catch {
+    return {
+      success: false,
+      status: 502,
+      error: `Invalid upstream response (status ${resp.status})`
+    };
+  }
+
+  if (!resp.ok) {
+    const errMsg =
+      data?.error?.message ||
+      data?.error ||
+      data?.message ||
+      `Upstream HTTP ${resp.status}`;
+    log?.warn?.(`[chatSearch] upstream error provider=${provider} status=${resp.status}`);
+    return {
+      success: false,
+      status: resp.status,
+      error: typeof errMsg === "string" ? errMsg : JSON.stringify(errMsg)
+    };
+  }
+
+  const { text, citations, tokens } = cfg.extractAnswer(data);
+  const retrievedAt = new Date().toISOString();
+  const limited = (citations || []).slice(0, limit);
+  const results = limited.map((c, i) => toResult(c, i, provider, retrievedAt));
+
+  return {
+    success: true,
+    status: 200,
+    data: {
+      provider,
+      query,
+      results,
+      answer: { source: provider, text: text || "", model: useModel },
+      usage: { queries_used: 1, search_cost_usd: 0, llm_tokens: tokens || 0 },
+      metrics: {
+        response_time_ms: Date.now() - startTime,
+        upstream_latency_ms: upstreamLatency,
+        total_results_available: null
+      },
+      errors: []
+    }
+  };
+}
+
+export { CHAT_SEARCH_CONFIG };
--- a/open-sse/handlers/search/index.js
+++ b/open-sse/handlers/search/index.js
@@ -0,0 +1,201 @@
+/**
+ * Search Dispatcher — routes /v1/search requests to dedicated search APIs
+ * or chat-based LLM search wrappers, with retry-friendly error envelope.
+ *
+ * Dependency map:
+ *   provider.searchConfig    → dedicated search API (callers + normalizers)
+ *   provider.searchViaChat   → wrap chat-completions (chatSearch.js)
+ */
+
+import { buildSearchRequest } from "./callers.js";
+import { normalizeSearchResponse } from "./normalizers.js";
+import { handleChatSearch } from "./chatSearch.js";
+
+const GLOBAL_TIMEOUT_MS = 15000;
+const NON_RETRIABLE = new Set([400, 401, 403, 404]);
+
+const CONTROL_CHAR_RE = /[\x00-\x08\x0B\x0C\x0E-\x1F\x7F]/;
+
+/** Normalize and validate query string. */
+function sanitizeQuery(query) {
+  if (CONTROL_CHAR_RE.test(query)) return { error: "Query contains invalid control characters" };
+  const clean = query.normalize("NFKC").trim().replace(/\s+/g, " ");
+  if (!clean) return { error: "Query is empty after normalization" };
+  return { clean };
+}
+
+// Strip non-ASCII chars from header values (HTTP headers must be ByteString).
+function sanitizeHeaders(headers) {
+  if (!headers) return headers;
+  const out = {};
+  for (const [k, v] of Object.entries(headers)) {
+    out[k] = typeof v === "string" ? v.replace(/[^\x00-\xFF]/g, "").trim() : v;
+  }
+  return out;
+}
+
+/** Build a JSON Response wrapper used by the auth layer. */
+function jsonResponse(payload, status = 200) {
+  return new Response(JSON.stringify(payload), {
+    status,
+    headers: { "Content-Type": "application/json", "Access-Control-Allow-Origin": "*" }
+  });
+}
+
+/** Wrap an error result with a Response object so the auth wrapper can return it directly. */
+function errorResult(status, error) {
+  return {
+    success: false,
+    status,
+    error,
+    response: jsonResponse({ error: { message: error, code: status } }, status)
+  };
+}
+
+/** Wrap a success payload. */
+function successResult(data) {
+  return { success: true, data, response: jsonResponse(data, 200) };
+}
+
+/**
+ * Run a single dedicated search provider attempt.
+ * @returns {Promise<{success:boolean, status?:number, error?:string, data?:object}>}
+ */
+async function tryDedicatedProvider({ provider, providerConfig, body, credentials, log, globalStartTime }) {
+  const startTime = Date.now();
+  const token = credentials?.apiKey || credentials?.accessToken || undefined;
+
+  if (providerConfig.authType !== "none" && !token) {
+    return { success: false, status: 401, error: `No credentials for provider: ${provider.id}` };
+  }
+
+  const params = {
+    query: body.query,
+    searchType: body.search_type || (providerConfig.searchTypes?.[0] || "web"),
+    maxResults: Math.min(body.max_results || providerConfig.defaultMaxResults || 5, providerConfig.maxMaxResults || 100),
+    token,
+    country: body.country,
+    language: body.language,
+    timeRange: body.time_range,
+    offset: body.offset,
+    domainFilter: body.domain_filter,
+    contentOptions: body.content_options,
+    providerOptions: body.provider_options,
+    providerSpecificData: credentials?.providerSpecificData
+  };
+
+  let url, init;
+  try {
+    ({ url, init } = buildSearchRequest({ id: provider.id, ...providerConfig }, params));
+  } catch (err) {
+    return { success: false, status: 400, error: err?.message || `Invalid request for ${provider.id}` };
+  }
+
+  // Timeout = min(provider timeout, remaining global)
+  const remaining = GLOBAL_TIMEOUT_MS - (Date.now() - globalStartTime);
+  const timeout = Math.min(providerConfig.timeoutMs || 10000, Math.max(remaining, 1000));
+  const controller = new AbortController();
+  const timer = setTimeout(() => controller.abort(), timeout);
+
+  log?.info?.("SEARCH", `${provider.id} | "${params.query.slice(0, 80)}" | type=${params.searchType}`);
+
+  try {
+    const resp = await fetch(url, { ...init, headers: sanitizeHeaders(init.headers), signal: controller.signal });
+    clearTimeout(timer);
+    if (!resp.ok) {
+      const errText = await resp.text().catch(() => "");
+      log?.error?.("SEARCH", `${provider.id} ${resp.status}: ${errText.slice(0, 200)}`);
+      return { success: false, status: resp.status, error: `${provider.id} returned ${resp.status}: ${errText.slice(0, 200)}` };
+    }
+    const data = await resp.json();
+    const normalized = normalizeSearchResponse(provider.id, data, params.query, params.searchType);
+    const results = normalized.results.slice(0, params.maxResults);
+    const duration = Date.now() - startTime;
+
+    return {
+      success: true,
+      data: {
+        provider: provider.id,
+        query: params.query,
+        results,
+        answer: null,
+        usage: { queries_used: 1, search_cost_usd: providerConfig.costPerQuery || 0 },
+        metrics: { response_time_ms: duration, upstream_latency_ms: duration, total_results_available: normalized.totalResults },
+        errors: []
+      }
+    };
+  } catch (err) {
+    clearTimeout(timer);
+    const isTimeout = err.name === "AbortError";
+    const status = isTimeout ? 504 : 502;
+    log?.error?.("SEARCH", `${provider.id} ${isTimeout ? "timeout" : "error"}: ${err.message}`);
+    return { success: false, status, error: `${provider.id} ${isTimeout ? "timeout" : "error"}: ${err.message}` };
+  }
+}
+
+/**
+ * Core search handler. Dispatches to dedicated API or chat-based LLM.
+ * Same calling convention as handleEmbeddingsCore: returns `{success, response, status?, error?}`.
+ *
+ * @param {object}   options
+ * @param {object}   options.body            Sanitized body from auth wrapper
+ * @param {object}   options.provider        Provider entry from AI_PROVIDERS
+ * @param {object}   [options.providerConfig] Provider's searchConfig (if dedicated)
+ * @param {object|null} options.credentials  Provider credentials
+ * @param {object}   [options.log]           Logger
+ */
+export async function handleSearchCore({ body, provider, providerConfig, credentials, log }) {
+  const globalStartTime = Date.now();
+
+  // 1. Sanitize query
+  const { clean, error: sanitizeError } = sanitizeQuery(body.query || "");
+  if (sanitizeError) return errorResult(400, sanitizeError);
+  const normalizedBody = { ...body, query: clean };
+
+  // 2. Route: dedicated search API takes priority over chat-based
+  let result;
+  if (providerConfig) {
+    result = await tryDedicatedProvider({
+      provider,
+      providerConfig,
+      body: normalizedBody,
+      credentials,
+      log,
+      globalStartTime
+    });
+  } else if (provider.searchViaChat) {
+    result = await handleChatSearch({
+      provider: provider.id,
+      query: clean,
+      maxResults: normalizedBody.max_results,
+      model: provider.searchViaChat.defaultModel,
+      credentials,
+      log
+    });
+  } else {
+    return errorResult(400, `Provider ${provider.id} does not support web search`);
+  }
+
+  if (result.success) return successResult(result.data);
+
+  // 3. Failover within global timeout for retriable errors
+  if (
+    !NON_RETRIABLE.has(result.status || 0) &&
+    Date.now() - globalStartTime < GLOBAL_TIMEOUT_MS &&
+    provider.searchViaChat &&
+    providerConfig
+  ) {
+    log?.warn?.("SEARCH", `${provider.id} dedicated failed (${result.status}), falling back to chat-based search`);
+    const fallback = await handleChatSearch({
+      provider: provider.id,
+      query: clean,
+      maxResults: normalizedBody.max_results,
+      model: provider.searchViaChat.defaultModel,
+      credentials,
+      log
+    });
+    if (fallback.success) return successResult(fallback.data);
+  }
+
+  return errorResult(result.status || 502, result.error || "Search failed");
+}
--- a/open-sse/handlers/search/normalizers.js
+++ b/open-sse/handlers/search/normalizers.js
@@ -0,0 +1,223 @@
+/**
+ * Search Response Normalizers
+ *
+ * Ported from OmniRoute open-sse/handlers/search.ts.
+ * Each normalizer maps a provider-specific response into the unified SearchResult shape.
+ */
+
+/** Build a unified SearchResult object. */
+function makeResult(providerId, item, idx, now) {
+  const url = item.url || "";
+  return {
+    title: item.title || "",
+    url,
+    display_url: url ? url.replace(/^https?:\/\/(www\.)?/, "").split("?")[0] : undefined,
+    snippet: item.snippet || "",
+    position: idx + 1,
+    score: typeof item.score === "number" ? Math.min(1, Math.max(0, item.score)) : null,
+    published_at: item.published_at || null,
+    favicon_url: item.favicon_url || null,
+    content: item.full_text
+      ? { format: item.text_format || "text", text: item.full_text, length: item.full_text.length }
+      : null,
+    metadata: {
+      author: item.author || null,
+      language: null,
+      source_type: item.source_type || null,
+      image_url: item.image_url || null,
+    },
+    citation: { provider: providerId, retrieved_at: now, rank: idx + 1 },
+    provider_raw: null,
+  };
+}
+
+function normalizeSerper(data, _query, searchType) {
+  const now = new Date().toISOString();
+  const items = searchType === "news" ? data.news : data.organic;
+  if (!Array.isArray(items)) return { results: [], totalResults: null };
+  const results = items.map((item, idx) =>
+    makeResult("serper", { title: item.title, url: item.link, snippet: item.snippet || item.description, published_at: item.date }, idx, now)
+  );
+  const total = data.searchParameters?.totalResults;
+  return { results, totalResults: typeof total === "number" ? total : null };
+}
+
+function normalizeBrave(data, _query, searchType) {
+  const now = new Date().toISOString();
+  const container = searchType === "news" ? data.news || data : data.web;
+  const items = container?.results;
+  if (!Array.isArray(items)) return { results: [], totalResults: null };
+  const results = items.map((item, idx) =>
+    makeResult("brave-search", {
+      title: item.title,
+      url: item.url,
+      snippet: item.description,
+      published_at: item.page_age || item.age,
+      favicon_url: item.meta_url?.favicon || item.favicon,
+    }, idx, now)
+  );
+  return { results, totalResults: container?.totalCount ?? null };
+}
+
+function normalizePerplexity(data, _query, _searchType) {
+  const now = new Date().toISOString();
+  const items = data.results;
+  if (!Array.isArray(items)) return { results: [], totalResults: null };
+  const results = items.map((item, idx) =>
+    makeResult("perplexity", { title: item.title, url: item.url, snippet: item.snippet, published_at: item.date || item.last_updated }, idx, now)
+  );
+  return { results, totalResults: results.length };
+}
+
+function normalizeExa(data, _query, _searchType) {
+  const now = new Date().toISOString();
+  const items = data.results;
+  if (!Array.isArray(items)) return { results: [], totalResults: null };
+  const results = items.map((item, idx) =>
+    makeResult("exa", {
+      title: item.title,
+      url: item.url,
+      snippet: item.highlights?.[0] || item.text?.slice(0, 300) || "",
+      score: item.score,
+      published_at: item.publishedDate,
+      favicon_url: item.favicon,
+      author: item.author,
+      image_url: item.image,
+      full_text: item.text,
+      text_format: "text",
+    }, idx, now)
+  );
+  return { results, totalResults: results.length };
+}
+
+function normalizeTavily(data, _query, _searchType) {
+  const now = new Date().toISOString();
+  const items = data.results;
+  if (!Array.isArray(items)) return { results: [], totalResults: null };
+  const results = items.map((item, idx) =>
+    makeResult("tavily", {
+      title: item.title,
+      url: item.url,
+      snippet: item.content || "",
+      score: item.score,
+      published_at: item.published_date,
+      full_text: item.raw_content,
+      text_format: "text",
+    }, idx, now)
+  );
+  return { results, totalResults: results.length };
+}
+
+function normalizeGooglePse(data, _query, _searchType) {
+  const now = new Date().toISOString();
+  const items = Array.isArray(data.items) ? data.items : [];
+  const results = items.map((item, idx) =>
+    makeResult("google-pse", {
+      title: item.title,
+      url: item.link,
+      snippet: item.snippet,
+      image_url: item.pagemap?.cse_image?.[0]?.src || item.pagemap?.cse_thumbnail?.[0]?.src || item.pagemap?.metatags?.[0]?.["og:image"],
+    }, idx, now)
+  );
+  const raw = data.searchInformation?.totalResults ?? data.queries?.request?.[0]?.totalResults ?? null;
+  const total = typeof raw === "string" ? Number(raw) : raw;
+  return { results, totalResults: Number.isFinite(total) ? total : null };
+}
+
+function normalizeLinkup(data, _query, _searchType) {
+  const now = new Date().toISOString();
+  const items = Array.isArray(data.results) ? data.results : [];
+  const results = items.map((item, idx) =>
+    makeResult("linkup", {
+      title: item.name || item.title,
+      url: item.url,
+      snippet: item.content || item.snippet || "",
+      source_type: item.type || "web",
+      image_url: item.image_url || item.imageUrl || null,
+      full_text: item.content,
+      text_format: "text",
+    }, idx, now)
+  );
+  return { results, totalResults: results.length };
+}
+
+function normalizeSearchApi(data, _query, _searchType) {
+  const now = new Date().toISOString();
+  const items = Array.isArray(data.organic_results) ? data.organic_results : Array.isArray(data.top_stories) ? data.top_stories : [];
+  const results = items.map((item, idx) =>
+    makeResult("searchapi", {
+      title: item.title,
+      url: item.link,
+      snippet: item.snippet || item.description || "",
+      published_at: item.date || item.published_at,
+      favicon_url: item.favicon,
+      author: item.source || null,
+      image_url: item.thumbnail || null,
+    }, idx, now)
+  );
+  const raw = data.search_information?.total_results;
+  const total = typeof raw === "number" ? raw : typeof raw === "string" ? Number(raw) : null;
+  return { results, totalResults: Number.isFinite(total) ? total : results.length };
+}
+
+function normalizeYouCom(data, _query, searchType) {
+  const now = new Date().toISOString();
+  const container = data?.results && typeof data.results === "object" ? data.results : undefined;
+  const section = searchType === "news" ? container?.news || [] : container?.web || [];
+  const items = Array.isArray(section) ? section : [];
+  const results = items.map((item, idx) => {
+    const firstSnippet = Array.isArray(item.snippets) ? item.snippets.find((v) => typeof v === "string") : null;
+    const livecrawlText = typeof item.markdown === "string" ? item.markdown : typeof item.html === "string" ? item.html : undefined;
+    const livecrawlFormat = typeof item.markdown === "string" ? "markdown" : "html";
+    return makeResult("youcom", {
+      title: item.title,
+      url: item.url,
+      snippet: typeof firstSnippet === "string" ? firstSnippet : typeof item.description === "string" ? item.description : "",
+      published_at: item.page_age,
+      favicon_url: item.favicon_url,
+      image_url: item.thumbnail_url,
+      source_type: searchType,
+      full_text: livecrawlText,
+      text_format: livecrawlText ? livecrawlFormat : undefined,
+    }, idx, now);
+  });
+  return { results, totalResults: results.length };
+}
+
+function normalizeSearxng(data, _query, _searchType) {
+  const now = new Date().toISOString();
+  const items = Array.isArray(data.results) ? data.results : [];
+  const results = items.map((item, idx) =>
+    makeResult("searxng", {
+      title: item.title,
+      url: item.url,
+      snippet: item.content || item.snippet || "",
+      published_at: item.publishedDate || item.published_date || null,
+      source_type: Array.isArray(item.engines) ? item.engines.join(", ") : item.engine || item.category || null,
+      image_url: item.thumbnail || item.img_src || null,
+    }, idx, now)
+  );
+  return { results, totalResults: results.length };
+}
+
+const NORMALIZERS = {
+  "serper": normalizeSerper,
+  "brave-search": normalizeBrave,
+  "perplexity": normalizePerplexity,
+  "exa": normalizeExa,
+  "tavily": normalizeTavily,
+  "google-pse": normalizeGooglePse,
+  "linkup": normalizeLinkup,
+  "searchapi": normalizeSearchApi,
+  "youcom": normalizeYouCom,
+  "searxng": normalizeSearxng,
+};
+
+/**
+ * Dispatch to the appropriate normalizer based on providerId.
+ * @returns {{results: Array, totalResults: number|null}}
+ */
+export function normalizeSearchResponse(providerId, data, query, searchType) {
+  const fn = NORMALIZERS[providerId];
+  return fn ? fn(data, query, searchType) : { results: [], totalResults: null };
+}
--- a/open-sse/services/tokenRefresh.js
+++ b/open-sse/services/tokenRefresh.js
@@ -1,5 +1,6 @@
 import { PROVIDERS } from "../config/providers.js";
 import { OAUTH_ENDPOINTS, GITHUB_COPILOT, REFRESH_LEAD_MS } from "../config/appConstants.js";
+import { proxyAwareFetch } from "../utils/proxyFetch.js";

 // Default token expiry buffer (refresh if expires within 5 minutes)
 export const TOKEN_EXPIRY_BUFFER_MS = 5 * 60 * 1000;
@@ -242,7 +243,7 @@ export async function refreshCodexToken(refreshToken, log) {
 * Specialized refresh for Kiro (AWS CodeWhisperer) tokens
 * Supports both AWS SSO OIDC (Builder ID/IDC) and Social Auth (Google/GitHub)
 */
-export async function refreshKiroToken(refreshToken, providerSpecificData, log) {
+export async function refreshKiroToken(refreshToken, providerSpecificData, log, proxyOptions = null) {
  const authMethod = providerSpecificData?.authMethod;
  const clientId = providerSpecificData?.clientId;
  const clientSecret = providerSpecificData?.clientSecret;
@@ -256,7 +257,7 @@ export async function refreshKiroToken(refreshToken, providerSpecificData, log)
      ? `https://oidc.${region}.amazonaws.com/token`
      : "https://oidc.us-east-1.amazonaws.com/token";

-    const response = await fetch(endpoint, {
+    const response = await proxyAwareFetch(endpoint, {
      method: "POST",
      headers: {
        "Content-Type": "application/json",
@@ -268,7 +269,7 @@ export async function refreshKiroToken(refreshToken, providerSpecificData, log)
        refreshToken: refreshToken,
        grantType: "refresh_token",
      }),
-    });
+    }, proxyOptions);

    if (!response.ok) {
      const errorText = await response.text();
@@ -294,7 +295,7 @@ export async function refreshKiroToken(refreshToken, providerSpecificData, log)
  }

  // Social Auth (Google/GitHub) - use Kiro's refresh endpoint
-  const response = await fetch(PROVIDERS.kiro.tokenUrl, {
+  const response = await proxyAwareFetch(PROVIDERS.kiro.tokenUrl, {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
@@ -304,7 +305,7 @@ export async function refreshKiroToken(refreshToken, providerSpecificData, log)
    body: JSON.stringify({
      refreshToken: refreshToken,
    }),
-  });
+  }, proxyOptions);

  if (!response.ok) {
    const errorText = await response.text();
--- a/open-sse/services/usage.js
+++ b/open-sse/services/usage.js
@@ -3,6 +3,7 @@
 */

 import { CLIENT_METADATA, getPlatformUserAgent } from "../config/appConstants.js";
+import { proxyAwareFetch } from "../utils/proxyFetch.js";

 // GitHub API config
 const GITHUB_CONFIG = {
@@ -38,22 +39,22 @@ const CLAUDE_CONFIG = {
 * @param {Object} connection - Provider connection with accessToken
 * @returns {Object} Usage data with quotas
 */
-export async function getUsageForProvider(connection) {
+export async function getUsageForProvider(connection, proxyOptions = null) {
  const { provider, accessToken, providerSpecificData } = connection;

  switch (provider) {
    case "github":
-      return await getGitHubUsage(accessToken, providerSpecificData);
+      return await getGitHubUsage(accessToken, providerSpecificData, proxyOptions);
    case "gemini-cli":
-      return await getGeminiUsage(accessToken);
+      return await getGeminiUsage(accessToken, proxyOptions);
    case "antigravity":
-      return await getAntigravityUsage(accessToken);
+      return await getAntigravityUsage(accessToken, providerSpecificData, proxyOptions);
    case "claude":
-      return await getClaudeUsage(accessToken);
+      return await getClaudeUsage(accessToken, proxyOptions);
    case "codex":
-      return await getCodexUsage(accessToken);
+      return await getCodexUsage(accessToken, proxyOptions);
    case "kiro":
-      return await getKiroUsage(accessToken, providerSpecificData);
+      return await getKiroUsage(accessToken, providerSpecificData, proxyOptions);
    case "qwen":
      return await getQwenUsage(accessToken, providerSpecificData);
    case "iflow":
@@ -103,14 +104,14 @@ function parseResetTime(resetValue) {
 * GitHub Copilot Usage
 * Uses GitHub accessToken (not copilotToken) to call copilot_internal/user API
 */
-async function getGitHubUsage(accessToken, providerSpecificData) {
+async function getGitHubUsage(accessToken, providerSpecificData, proxyOptions = null) {
  try {
    if (!accessToken) {
      throw new Error("No GitHub access token available. Please re-authorize the connection.");
    }

    // copilot_internal/user API requires GitHub OAuth token, not copilotToken
-    const response = await fetch("https://api.github.com/copilot_internal/user", {
+    const response = await proxyAwareFetch("https://api.github.com/copilot_internal/user", {
      headers: {
        "Authorization": `token ${accessToken}`,
        "Accept": "application/json",
@@ -119,7 +120,7 @@ async function getGitHubUsage(accessToken, providerSpecificData) {
        "Editor-Version": "vscode/1.100.0",
        "Editor-Plugin-Version": "copilot-chat/0.26.7",
      },
-    });
+    }, proxyOptions);

    if (!response.ok) {
      const error = await response.text();
@@ -189,18 +190,19 @@ function formatGitHubQuotaSnapshot(quota) {
 /**
 * Gemini CLI Usage (Google Cloud)
 */
-async function getGeminiUsage(accessToken) {
+async function getGeminiUsage(accessToken, proxyOptions = null) {
  try {
    // Gemini CLI uses Google Cloud quotas
    // Try to get quota info from Cloud Resource Manager
-    const response = await fetch(
+    const response = await proxyAwareFetch(
      "https://cloudresourcemanager.googleapis.com/v1/projects?filter=lifecycleState:ACTIVE",
      {
        headers: {
          Authorization: `Bearer ${accessToken}`,
          Accept: "application/json",
        },
-      }
+      },
+      proxyOptions
    );

    if (!response.ok) {
@@ -217,10 +219,10 @@ async function getGeminiUsage(accessToken) {
 /**
 * Antigravity Usage - Fetch quota from Google Cloud Code API
 */
-async function getAntigravityUsage(accessToken, providerSpecificData) {
+async function getAntigravityUsage(accessToken, providerSpecificData, proxyOptions = null) {
  try {
    // Fetch subscription info once — reuse for both projectId and plan
-    const subscriptionInfo = await getAntigravitySubscriptionInfo(accessToken);
+    const subscriptionInfo = await getAntigravitySubscriptionInfo(accessToken, proxyOptions);
    const projectId = subscriptionInfo?.cloudaicompanionProject || null;

    // Fetch quota data with timeout
@@ -229,7 +231,7 @@ async function getAntigravityUsage(accessToken, providerSpecificData) {

    let response;
    try {
-      response = await fetch(ANTIGRAVITY_CONFIG.quotaApiUrl, {
+      response = await proxyAwareFetch(ANTIGRAVITY_CONFIG.quotaApiUrl, {
        method: "POST",
        headers: {
          "Authorization": `Bearer ${accessToken}`,
@@ -243,7 +245,7 @@ async function getAntigravityUsage(accessToken, providerSpecificData) {
          ...(projectId ? { project: projectId } : {})
        }),
        signal: controller.signal,
-      });
+      }, proxyOptions);
    } finally {
      clearTimeout(timeoutId);
    }
@@ -338,11 +340,11 @@ async function getAntigravityProjectId(accessToken) {
 /**
 * Get Antigravity subscription info
 */
-async function getAntigravitySubscriptionInfo(accessToken) {
+async function getAntigravitySubscriptionInfo(accessToken, proxyOptions = null) {
  const controller = new AbortController();
  const timeoutId = setTimeout(() => controller.abort(), 10000); // 10s timeout
  try {
-    const response = await fetch(ANTIGRAVITY_CONFIG.loadProjectApiUrl, {
+    const response = await proxyAwareFetch(ANTIGRAVITY_CONFIG.loadProjectApiUrl, {
      method: "POST",
      headers: {
        "Authorization": `Bearer ${accessToken}`,
@@ -352,7 +354,7 @@ async function getAntigravitySubscriptionInfo(accessToken) {
      },
      body: JSON.stringify({ metadata: CLIENT_METADATA, mode: 1 }),
      signal: controller.signal,
-    });
+    }, proxyOptions);

    if (!response.ok) return null;
    return await response.json();
@@ -367,17 +369,17 @@ async function getAntigravitySubscriptionInfo(accessToken) {
 /**
 * Claude Usage - Primary: OAuth endpoint, Fallback: legacy settings/org endpoint
 */
-async function getClaudeUsage(accessToken) {
+async function getClaudeUsage(accessToken, proxyOptions = null) {
  try {
    // Primary: OAuth usage endpoint (Claude Code consumer OAuth tokens)
-    const oauthResponse = await fetch(CLAUDE_CONFIG.oauthUsageUrl, {
+    const oauthResponse = await proxyAwareFetch(CLAUDE_CONFIG.oauthUsageUrl, {
      method: "GET",
      headers: {
        "Authorization": `Bearer ${accessToken}`,
        "anthropic-beta": "oauth-2025-04-20",
        "anthropic-version": CLAUDE_CONFIG.apiVersion,
      },
-    });
+    }, proxyOptions);

    if (oauthResponse.ok) {
      const data = await oauthResponse.json();
@@ -425,7 +427,7 @@ async function getClaudeUsage(accessToken) {

    // Fallback: legacy settings + org usage endpoint
    console.warn(`[Claude Usage] OAuth endpoint returned ${oauthResponse.status}, falling back to legacy`);
-    return await getClaudeUsageLegacy(accessToken);
+    return await getClaudeUsageLegacy(accessToken, proxyOptions);
  } catch (error) {
    return { message: `Claude connected. Unable to fetch usage: ${error.message}` };
  }
@@ -434,21 +436,21 @@ async function getClaudeUsage(accessToken) {
 /**
 * Legacy Claude usage for API key / org admin users
 */
-async function getClaudeUsageLegacy(accessToken) {
+async function getClaudeUsageLegacy(accessToken, proxyOptions = null) {
  try {
-    const settingsResponse = await fetch(CLAUDE_CONFIG.settingsUrl, {
+    const settingsResponse = await proxyAwareFetch(CLAUDE_CONFIG.settingsUrl, {
      method: "GET",
      headers: {
        "Authorization": `Bearer ${accessToken}`,
        "anthropic-version": CLAUDE_CONFIG.apiVersion,
      },
-    });
+    }, proxyOptions);

    if (settingsResponse.ok) {
      const settings = await settingsResponse.json();

      if (settings.organization_id) {
-        const usageResponse = await fetch(
+        const usageResponse = await proxyAwareFetch(
          CLAUDE_CONFIG.usageUrl.replace("{org_id}", settings.organization_id),
          {
            method: "GET",
@@ -456,7 +458,8 @@ async function getClaudeUsageLegacy(accessToken) {
              "Authorization": `Bearer ${accessToken}`,
              "anthropic-version": CLAUDE_CONFIG.apiVersion,
            },
-          }
+          },
+          proxyOptions
        );

        if (usageResponse.ok) {
@@ -485,15 +488,15 @@ async function getClaudeUsageLegacy(accessToken) {
 /**
 * Codex (OpenAI) Usage - Fetch from ChatGPT backend API
 */
-async function getCodexUsage(accessToken) {
+async function getCodexUsage(accessToken, proxyOptions = null) {
  try {
-    const response = await fetch(CODEX_CONFIG.usageUrl, {
+    const response = await proxyAwareFetch(CODEX_CONFIG.usageUrl, {
      method: "GET",
      headers: {
        "Authorization": `Bearer ${accessToken}`,
        "Accept": "application/json",
      },
-    });
+    }, proxyOptions);

    if (!response.ok) {
      return { message: `Codex connected. Usage API temporarily unavailable (${response.status}).` };
@@ -577,7 +580,7 @@ function parseKiroQuotaData(data) {
  };
 }

-async function getKiroUsage(accessToken, providerSpecificData) {
+async function getKiroUsage(accessToken, providerSpecificData, proxyOptions = null) {
  // Default profileArn fallback
  const DEFAULT_PROFILE_ARN = "arn:aws:codewhisperer:us-east-1:638616132270:profile/AAAACCCCXXXX";
  const profileArn = providerSpecificData?.profileArn || DEFAULT_PROFILE_ARN;
@@ -593,7 +596,7 @@ async function getKiroUsage(accessToken, providerSpecificData) {
  const attempts = [
    {
      name: "codewhisperer-get",
-      run: async () => fetch(
+      run: async () => proxyAwareFetch(
        `https://codewhisperer.us-east-1.amazonaws.com/getUsageLimits?${getUsageParams.toString()}`,
        {
          method: "GET",
@@ -604,11 +607,12 @@ async function getKiroUsage(accessToken, providerSpecificData) {
            "user-agent": "aws-sdk-js/1.0.0 KiroIDE",
          },
        },
+        proxyOptions
      ),
    },
    {
      name: "codewhisperer-post",
-      run: async () => fetch("https://codewhisperer.us-east-1.amazonaws.com", {
+      run: async () => proxyAwareFetch("https://codewhisperer.us-east-1.amazonaws.com", {
        method: "POST",
        headers: {
          "Authorization": `Bearer ${accessToken}`,
@@ -621,7 +625,7 @@ async function getKiroUsage(accessToken, providerSpecificData) {
          profileArn,
          resourceType: "AGENTIC_REQUEST",
        }),
-      }),
+      }, proxyOptions),
    },
    {
      name: "q-get",
@@ -631,13 +635,13 @@ async function getKiroUsage(accessToken, providerSpecificData) {
          profileArn,
          resourceType: "AGENTIC_REQUEST",
        });
-        return fetch(`https://q.us-east-1.amazonaws.com/getUsageLimits?${params}`, {
+        return proxyAwareFetch(`https://q.us-east-1.amazonaws.com/getUsageLimits?${params}`, {
          method: "GET",
          headers: {
            "Authorization": `Bearer ${accessToken}`,
            "Accept": "application/json",
          },
-        });
+        }, proxyOptions);
      },
    },
  ];
--- a/public/providers/azure.png
+++ b/public/providers/azure.png
--- a/public/providers/blackbox.png
+++ b/public/providers/blackbox.png
--- a/public/providers/brave-search.png
+++ b/public/providers/brave-search.png
--- a/public/providers/exa.png
+++ b/public/providers/exa.png
--- a/public/providers/firecrawl.png
+++ b/public/providers/firecrawl.png
--- a/public/providers/google-pse.png
+++ b/public/providers/google-pse.png
--- a/public/providers/grok-web.png
+++ b/public/providers/grok-web.png
--- a/public/providers/jina-reader.png
+++ b/public/providers/jina-reader.png
--- a/public/providers/linkup.png
+++ b/public/providers/linkup.png
--- a/public/providers/perplexity-web.png
+++ b/public/providers/perplexity-web.png
--- a/public/providers/searchapi.png
+++ b/public/providers/searchapi.png
--- a/public/providers/searxng.png
+++ b/public/providers/searxng.png
--- a/public/providers/serper.png
+++ b/public/providers/serper.png
--- a/public/providers/tavily.png
+++ b/public/providers/tavily.png
--- a/public/providers/youcom.png
+++ b/public/providers/youcom.png
--- a/src/app/(dashboard)/dashboard/combos/page.js
+++ b/src/app/(dashboard)/dashboard/combos/page.js
@@ -339,7 +339,7 @@ function ModelItem({ index, model, isFirst, isLast, onEdit, onMoveUp, onMoveDown
  );
 }

-function ComboFormModal({ isOpen, combo, onClose, onSave, activeProviders }) {
+function ComboFormModal({ isOpen, combo, onClose, onSave, activeProviders, kindFilter = null }) {
  // Initialize state with combo values - key prop on parent handles reset on remount
  const [name, setName] = useState(combo?.name || "");
  const [models, setModels] = useState(combo?.models || []);
@@ -504,6 +504,7 @@ function ComboFormModal({ isOpen, combo, onClose, onSave, activeProviders }) {
        activeProviders={activeProviders}
        modelAliases={modelAliases}
        title="Add Model to Combo"
+        kindFilter={kindFilter}
      />
    </>
  );
--- a/src/app/(dashboard)/dashboard/media-providers/[kind]/[id]/page.js
+++ b/src/app/(dashboard)/dashboard/media-providers/[kind]/[id]/page.js
@@ -1446,10 +1446,15 @@ export default function MediaProviderDetailPage() {
        />
      )}

-      {/* Provider Info — config-driven, only for providers with searchConfig/fetchConfig */}
-      {!isCustom && (provider.searchConfig || provider.fetchConfig) && (
+      {/* Provider Info — config-driven, supports searchConfig, fetchConfig, searchViaChat */}
+      {!isCustom && (provider.searchConfig || provider.fetchConfig || provider.searchViaChat) && (
        <ProviderInfoCard
-          config={kind === "webFetch" ? provider.fetchConfig : provider.searchConfig}
+          config={
+            kind === "webFetch"
+              ? provider.fetchConfig
+              : provider.searchConfig || { mode: "chat-completions", defaultModel: provider.searchViaChat?.defaultModel, costPerQuery: 0 }
+          }
+          provider={provider}
          title={`${kindConfig.label} Config`}
        />
      )}
--- a/src/app/(dashboard)/dashboard/media-providers/[kind]/page.js
+++ b/src/app/(dashboard)/dashboard/media-providers/[kind]/page.js
@@ -1,6 +1,6 @@
 "use client";

-import { useParams, notFound } from "next/navigation";
+import { useParams, notFound, useRouter } from "next/navigation";
 import Link from "next/link";
 import { useEffect, useState } from "react";
 import { Card, Badge, Button, AddCustomEmbeddingModal } from "@/shared/components";
@@ -72,10 +72,18 @@ function MediaProviderCard({ provider, kind, connections, isCustom }) {

 export default function MediaProviderKindPage() {
  const { kind } = useParams();
+  const router = useRouter();
  const [connections, setConnections] = useState([]);
  const [customNodes, setCustomNodes] = useState([]);
  const [showAddCustomEmbedding, setShowAddCustomEmbedding] = useState(false);

+  // webSearch/webFetch listing pages are merged into /web
+  useEffect(() => {
+    if (kind === "webSearch" || kind === "webFetch") {
+      router.replace("/dashboard/media-providers/web");
+    }
+  }, [kind, router]);
+
  const kindConfig = MEDIA_PROVIDER_KINDS.find((k) => k.id === kind);
  const isEmbedding = kind === "embedding";

--- a/src/app/(dashboard)/dashboard/media-providers/web/combo/[id]/page.js
+++ b/src/app/(dashboard)/dashboard/media-providers/web/combo/[id]/page.js
@@ -0,0 +1,346 @@
+"use client";
+
+import { useParams, notFound, useRouter } from "next/navigation";
+import { useEffect, useState } from "react";
+import Link from "next/link";
+import { Card, Button, Input, Toggle, Modal } from "@/shared/components";
+import ProviderIcon from "@/shared/components/ProviderIcon";
+import { AI_PROVIDERS, getProvidersByKind } from "@/shared/constants/providers";
+
+const VALID_NAME_REGEX = /^[a-zA-Z0-9_.\-]+$/;
+
+const KIND_LABELS = {
+  webSearch: "Web Search",
+  webFetch: "Web Fetch",
+};
+
+const EXAMPLE_PATHS = {
+  webSearch: "/v1/search",
+  webFetch: "/v1/web/fetch",
+};
+
+const EXAMPLE_BODIES = {
+  webSearch: (comboName) => ({ model: comboName, query: "What is the latest news about AI?", search_type: "web", max_results: 5 }),
+  webFetch: (comboName) => ({ model: comboName, url: "https://example.com", format: "markdown" }),
+};
+
+function ProviderPickerModal({ isOpen, onClose, onPick, kind, currentIds, connections }) {
+  // Only show providers with at least one usable connection (active/success) or noAuth
+  const usableIds = new Set(
+    (connections || [])
+      .filter((c) => {
+        if (c.isActive === false) return false;
+        const s = c.testStatus;
+        return s === "active" || s === "success" || s === "unavailable";
+      })
+      .map((c) => c.provider)
+  );
+  const all = kind ? getProvidersByKind(kind) : [];
+  const providers = all.filter((p) => p.noAuth || usableIds.has(p.id));
+  return (
+    <Modal isOpen={isOpen} onClose={onClose} title={`Add ${KIND_LABELS[kind] || ""} Provider`} size="md">
+      {providers.length === 0 ? (
+        <div className="text-center py-6 text-sm text-text-muted">
+          No connected providers available. Add a connection first in the {KIND_LABELS[kind]} section.
+        </div>
+      ) : (
+      <div className="grid grid-cols-2 sm:grid-cols-3 gap-2 max-h-[400px] overflow-y-auto">
+        {providers.map((p) => {
+          const already = currentIds.includes(p.id);
+          return (
+            <button
+              key={p.id}
+              disabled={already}
+              onClick={() => { onPick(p.id); onClose(); }}
+              className={`flex items-center gap-2 p-2 rounded-lg border transition-colors ${
+                already
+                  ? "border-border opacity-40 cursor-not-allowed"
+                  : "border-border hover:border-primary/50 hover:bg-primary/5 cursor-pointer"
+              }`}
+            >
+              <ProviderIcon
+                src={`/providers/${p.id}.png`}
+                alt={p.name}
+                size={24}
+                className="object-contain rounded shrink-0"
+                fallbackText={p.textIcon || p.id.slice(0, 2).toUpperCase()}
+                fallbackColor={p.color}
+              />
+              <span className="text-xs font-medium truncate text-left">{p.name}</span>
+              {already && <span className="text-[9px] text-text-muted ml-auto">added</span>}
+            </button>
+          );
+        })}
+      </div>
+      )}
+    </Modal>
+  );
+}
+
+export default function ComboDetailPage() {
+  const { id } = useParams();
+  const router = useRouter();
+  const [combo, setCombo] = useState(null);
+  const [loading, setLoading] = useState(true);
+  const [name, setName] = useState("");
+  const [nameError, setNameError] = useState("");
+  const [providers, setProviders] = useState([]);
+  const [roundRobin, setRoundRobin] = useState(false);
+  const [showPicker, setShowPicker] = useState(false);
+  const [logs, setLogs] = useState([]);
+  const [testing, setTesting] = useState(false);
+  const [testResult, setTestResult] = useState("");
+  const [apiKey, setApiKey] = useState("");
+  const [connections, setConnections] = useState([]);
+
+  const fetchAll = async () => {
+    try {
+      const [comboRes, settingsRes, logsRes, keysRes, connsRes] = await Promise.all([
+        fetch(`/api/combos/${id}`, { cache: "no-store" }),
+        fetch("/api/settings", { cache: "no-store" }),
+        fetch("/api/usage/logs", { cache: "no-store" }),
+        fetch("/api/keys", { cache: "no-store" }),
+        fetch("/api/providers", { cache: "no-store" }),
+      ]);
+      if (keysRes.ok) {
+        const k = await keysRes.json();
+        setApiKey((k.keys || []).find((x) => x.isActive !== false)?.key || "");
+      }
+      if (connsRes.ok) setConnections((await connsRes.json()).connections || []);
+      if (!comboRes.ok) { setCombo(null); setLoading(false); return; }
+      const c = await comboRes.json();
+      setCombo(c);
+      setName(c.name);
+      setProviders(c.models || []);
+      const s = settingsRes.ok ? await settingsRes.json() : {};
+      setRoundRobin(s.comboStrategies?.[c.name]?.fallbackStrategy === "round-robin");
+      const allLogs = logsRes.ok ? await logsRes.json() : [];
+      setLogs(allLogs.filter((l) => typeof l === "string" && l.includes(c.name)).slice(0, 50));
+    } catch { /* noop */ }
+    setLoading(false);
+  };
+
+  // eslint-disable-next-line react-hooks/set-state-in-effect
+  useEffect(() => { fetchAll(); }, [id]); // eslint-disable-line react-hooks/exhaustive-deps
+
+  const validateName = (v) => {
+    if (!v.trim()) { setNameError("Name is required"); return false; }
+    if (!VALID_NAME_REGEX.test(v)) { setNameError("Only letters, numbers, -, _ and ."); return false; }
+    setNameError("");
+    return true;
+  };
+
+  const saveCombo = async (patch) => {
+    const res = await fetch(`/api/combos/${id}`, {
+      method: "PUT",
+      headers: { "Content-Type": "application/json" },
+      body: JSON.stringify(patch),
+    });
+    if (!res.ok) { const err = await res.json(); alert(err.error || "Failed to save"); return false; }
+    return true;
+  };
+
+  const handleSaveName = async () => {
+    if (!validateName(name)) return;
+    if (name === combo.name) return;
+    const ok = await saveCombo({ name });
+    if (ok) await fetchAll();
+  };
+
+  const handleAddProvider = async (providerId) => {
+    const next = [...providers, providerId];
+    setProviders(next);
+    await saveCombo({ models: next });
+  };
+
+  const handleRemoveProvider = async (idx) => {
+    const next = providers.filter((_, i) => i !== idx);
+    setProviders(next);
+    await saveCombo({ models: next });
+  };
+
+  const handleMove = async (idx, dir) => {
+    const next = [...providers];
+    const swap = idx + dir;
+    if (swap < 0 || swap >= next.length) return;
+    [next[idx], next[swap]] = [next[swap], next[idx]];
+    setProviders(next);
+    await saveCombo({ models: next });
+  };
+
+  const handleToggleRoundRobin = async (enabled) => {
+    setRoundRobin(enabled);
+    const settingsRes = await fetch("/api/settings", { cache: "no-store" });
+    const s = settingsRes.ok ? await settingsRes.json() : {};
+    const updated = { ...(s.comboStrategies || {}) };
+    if (enabled) updated[combo.name] = { fallbackStrategy: "round-robin" };
+    else delete updated[combo.name];
+    await fetch("/api/settings", {
+      method: "PATCH",
+      headers: { "Content-Type": "application/json" },
+      body: JSON.stringify({ comboStrategies: updated }),
+    });
+  };
+
+  const handleDelete = async () => {
+    if (!confirm(`Delete combo "${combo.name}"?`)) return;
+    const res = await fetch(`/api/combos/${id}`, { method: "DELETE" });
+    if (res.ok) router.push("/dashboard/media-providers/web");
+  };
+
+  const handleTest = async () => {
+    setTesting(true);
+    setTestResult("");
+    try {
+      const path = EXAMPLE_PATHS[combo.kind];
+      const body = EXAMPLE_BODIES[combo.kind](combo.name);
+      const headers = { "Content-Type": "application/json" };
+      if (apiKey) headers["Authorization"] = `Bearer ${apiKey}`;
+      const res = await fetch(`/api${path}`, { method: "POST", headers, body: JSON.stringify(body) });
+      const data = await res.json().catch(() => ({}));
+      setTestResult(JSON.stringify(data, null, 2));
+    } catch (e) {
+      setTestResult(`Error: ${e.message}`);
+    }
+    setTesting(false);
+  };
+
+  if (loading) return <div className="text-text-muted text-sm">Loading...</div>;
+  if (!combo) return notFound();
+
+  const kindLabel = KIND_LABELS[combo.kind] || "Web";
+  const examplePath = EXAMPLE_PATHS[combo.kind];
+  const exampleBody = combo.kind ? EXAMPLE_BODIES[combo.kind](combo.name) : null;
+  const curlExample = examplePath
+    ? `curl -X POST http://localhost:20128${examplePath} \\\n  -H "Content-Type: application/json" \\\n  -H "Authorization: Bearer ${apiKey || "YOUR_KEY"}" \\\n  -d '${JSON.stringify(exampleBody)}'`
+    : "";
+
+  return (
+    <div className="flex flex-col gap-6">
+      {/* Header */}
+      <div className="flex items-center justify-between gap-4">
+        <div className="flex items-center gap-3 min-w-0">
+          <Link href="/dashboard/media-providers/web" className="text-text-muted hover:text-primary">
+            <span className="material-symbols-outlined">arrow_back</span>
+          </Link>
+          <div className="size-10 rounded-lg bg-primary/10 flex items-center justify-center">
+            <span className="material-symbols-outlined text-primary">layers</span>
+          </div>
+          <div className="min-w-0">
+            <p className="text-xs text-text-muted">{kindLabel} Combo</p>
+            <code className="text-lg font-semibold font-mono">{combo.name}</code>
+          </div>
+        </div>
+        <Button variant="outline" icon="delete" onClick={handleDelete} className="text-red-500 border-red-200 hover:bg-red-50">
+          Delete
+        </Button>
+      </div>
+
+      {/* Settings Card */}
+      <Card>
+        <h2 className="text-lg font-semibold mb-3">Settings</h2>
+        <div className="flex flex-col gap-4">
+          <div>
+            <Input label="Combo Name" value={name} onChange={(e) => { setName(e.target.value); validateName(e.target.value); }} onBlur={handleSaveName} error={nameError} />
+            <p className="text-[10px] text-text-muted mt-0.5">Only letters, numbers, -, _ and .</p>
+          </div>
+          <div className="flex items-center justify-between">
+            <div>
+              <p className="text-sm font-medium">Round Robin</p>
+              <p className="text-xs text-text-muted">Rotate providers across requests instead of strict fallback order.</p>
+            </div>
+            <Toggle checked={roundRobin} onChange={handleToggleRoundRobin} />
+          </div>
+        </div>
+      </Card>
+
+      {/* Providers Card */}
+      <Card>
+        <div className="flex items-center justify-between mb-3">
+          <div>
+            <h2 className="text-lg font-semibold">Providers</h2>
+            <p className="text-xs text-text-muted">Tried in order (top-down) or rotated when round-robin is on.</p>
+          </div>
+          <Button size="sm" icon="add" onClick={() => setShowPicker(true)}>Add Provider</Button>
+        </div>
+        {providers.length === 0 ? (
+          <div className="text-center py-6 border border-dashed border-border rounded-lg text-text-muted text-sm">
+            No providers yet.
+          </div>
+        ) : (
+          <div className="flex flex-col gap-2">
+            {providers.map((pid, idx) => {
+              const p = AI_PROVIDERS[pid];
+              return (
+                <div key={`${pid}-${idx}`} className="flex items-center gap-3 p-2 rounded-lg bg-black/[0.02] dark:bg-white/[0.02]">
+                  <span className="text-xs text-text-muted w-5 text-center">{idx + 1}</span>
+                  <ProviderIcon
+                    src={`/providers/${pid}.png`}
+                    alt={p?.name || pid}
+                    size={24}
+                    className="object-contain rounded shrink-0"
+                    fallbackText={p?.textIcon || pid.slice(0, 2).toUpperCase()}
+                    fallbackColor={p?.color}
+                  />
+                  <span className="text-sm font-medium flex-1 truncate">{p?.name || pid}</span>
+                  <div className="flex items-center gap-0.5">
+                    <button onClick={() => handleMove(idx, -1)} disabled={idx === 0} className={`p-1 rounded ${idx === 0 ? "text-text-muted/20" : "text-text-muted hover:text-primary hover:bg-black/5"}`} title="Move up">
+                      <span className="material-symbols-outlined text-[16px]">arrow_upward</span>
+                    </button>
+                    <button onClick={() => handleMove(idx, 1)} disabled={idx === providers.length - 1} className={`p-1 rounded ${idx === providers.length - 1 ? "text-text-muted/20" : "text-text-muted hover:text-primary hover:bg-black/5"}`} title="Move down">
+                      <span className="material-symbols-outlined text-[16px]">arrow_downward</span>
+                    </button>
+                    <button onClick={() => handleRemoveProvider(idx)} className="p-1 rounded text-text-muted hover:text-red-500 hover:bg-red-500/10" title="Remove">
+                      <span className="material-symbols-outlined text-[16px]">close</span>
+                    </button>
+                  </div>
+                </div>
+              );
+            })}
+          </div>
+        )}
+      </Card>
+
+      {/* Test Example Card */}
+      {combo.kind && (
+        <Card>
+          <div className="flex items-center justify-between mb-3">
+            <h2 className="text-lg font-semibold">Test Example</h2>
+            <Button size="sm" icon="play_arrow" onClick={handleTest} disabled={testing || providers.length === 0}>
+              {testing ? "Running..." : "Run"}
+            </Button>
+          </div>
+          <pre className="text-xs font-mono bg-black/[0.03] dark:bg-white/[0.03] p-3 rounded-lg overflow-x-auto whitespace-pre-wrap break-all">
+            {curlExample}
+          </pre>
+          {testResult && (
+            <pre className="mt-3 text-xs font-mono bg-black/[0.03] dark:bg-white/[0.03] p-3 rounded-lg overflow-auto max-h-[300px]">
+              {testResult}
+            </pre>
+          )}
+        </Card>
+      )}
+
+      {/* Usage Logs Card */}
+      <Card>
+        <h2 className="text-lg font-semibold mb-3">Usage Logs</h2>
+        {logs.length === 0 ? (
+          <p className="text-xs text-text-muted italic">No usage yet.</p>
+        ) : (
+          <pre className="text-[11px] font-mono bg-black/[0.03] dark:bg-white/[0.03] p-3 rounded-lg overflow-auto max-h-[400px] whitespace-pre-wrap">
+            {logs.join("\n")}
+          </pre>
+        )}
+      </Card>
+
+      <ProviderPickerModal
+        isOpen={showPicker}
+        onClose={() => setShowPicker(false)}
+        onPick={handleAddProvider}
+        kind={combo.kind}
+        currentIds={providers}
+        connections={connections}
+      />
+    </div>
+  );
+}
--- a/src/app/(dashboard)/dashboard/media-providers/web/page.js
+++ b/src/app/(dashboard)/dashboard/media-providers/web/page.js
@@ -0,0 +1,208 @@
+"use client";
+
+import Link from "next/link";
+import { useEffect, useState } from "react";
+import { useRouter } from "next/navigation";
+import { Card, Badge, Button } from "@/shared/components";
+import ProviderIcon from "@/shared/components/ProviderIcon";
+import { AI_PROVIDERS, getProvidersByKind } from "@/shared/constants/providers";
+
+function getEffectiveStatus(conn) {
+  const isCooldown = Object.entries(conn).some(
+    ([k, v]) => k.startsWith("modelLock_") && v && new Date(v).getTime() > Date.now()
+  );
+  return conn.testStatus === "unavailable" && !isCooldown ? "active" : conn.testStatus;
+}
+
+function ProviderCard({ provider, kind, connections }) {
+  const providerInfo = AI_PROVIDERS[provider.id];
+  const isNoAuth = !!providerInfo?.noAuth;
+  const providerConns = connections.filter((c) => c.provider === provider.id);
+  const connected = providerConns.filter((c) => { const s = getEffectiveStatus(c); return s === "active" || s === "success"; }).length;
+  const error = providerConns.filter((c) => { const s = getEffectiveStatus(c); return s === "error" || s === "expired" || s === "unavailable"; }).length;
+  const total = providerConns.length;
+  const allDisabled = total > 0 && providerConns.every((c) => c.isActive === false);
+
+  const renderStatus = () => {
+    if (isNoAuth) return <Badge variant="success" size="sm">Ready</Badge>;
+    if (allDisabled) return <Badge variant="default" size="sm">Disabled</Badge>;
+    if (total === 0) return <span className="text-xs text-text-muted">No connections</span>;
+    return (
+      <>
+        {connected > 0 && <Badge variant="success" size="sm" dot>{connected} Connected</Badge>}
+        {error > 0 && <Badge variant="error" size="sm" dot>{error} Error</Badge>}
+        {connected === 0 && error === 0 && <Badge variant="default" size="sm">{total} Added</Badge>}
+      </>
+    );
+  };
+
+  return (
+    <Link href={`/dashboard/media-providers/${kind}/${provider.id}`} className="group">
+      <Card padding="xs" className={`h-full hover:bg-black/[0.01] dark:hover:bg-white/[0.01] transition-colors cursor-pointer ${allDisabled ? "opacity-50" : ""}`}>
+        <div className="flex items-center gap-3">
+          <div
+            className="size-8 rounded-lg flex items-center justify-center shrink-0"
+            style={{ backgroundColor: `${provider.color?.length > 7 ? provider.color : (provider.color ?? "#888") + "15"}` }}
+          >
+            <ProviderIcon
+              src={`/providers/${provider.id}.png`}
+              alt={provider.name}
+              size={30}
+              className="object-contain rounded-lg max-w-[30px] max-h-[30px]"
+              fallbackText={provider.textIcon || provider.id.slice(0, 2).toUpperCase()}
+              fallbackColor={provider.color}
+            />
+          </div>
+          <div>
+            <h3 className="font-semibold text-sm">{provider.name}</h3>
+            <div className="flex items-center gap-2 mt-0.5 flex-wrap">{renderStatus()}</div>
+          </div>
+        </div>
+      </Card>
+    </Link>
+  );
+}
+
+function ComboList({ combos }) {
+  if (combos.length === 0) {
+    return <p className="text-xs text-text-muted italic">No combos yet.</p>;
+  }
+  return (
+    <div className="flex flex-col gap-2">
+      {combos.map((combo) => (
+        <Link key={combo.id} href={`/dashboard/media-providers/web/combo/${combo.id}`}>
+          <Card padding="xs" className="hover:bg-black/[0.02] dark:hover:bg-white/[0.02] transition-colors cursor-pointer">
+            <div className="flex items-center gap-3">
+              <span className="material-symbols-outlined text-primary text-[18px]">layers</span>
+              <code className="text-sm font-mono font-medium flex-1 truncate">{combo.name}</code>
+              {/* Provider icons preview */}
+              <div className="flex items-center gap-1 shrink-0">
+                {combo.models.slice(0, 6).map((pid, i) => {
+                  const p = AI_PROVIDERS[pid];
+                  return (
+                    <div key={`${pid}-${i}`} title={p?.name || pid} className="size-5 rounded flex items-center justify-center" style={{ backgroundColor: `${(p?.color ?? "#888")}15` }}>
+                      <ProviderIcon
+                        src={`/providers/${pid}.png`}
+                        alt={p?.name || pid}
+                        size={18}
+                        className="object-contain rounded max-w-[18px] max-h-[18px]"
+                        fallbackText={p?.textIcon || pid.slice(0, 2).toUpperCase()}
+                        fallbackColor={p?.color}
+                      />
+                    </div>
+                  );
+                })}
+                {combo.models.length > 6 && (
+                  <span className="text-[10px] text-text-muted ml-1">+{combo.models.length - 6}</span>
+                )}
+              </div>
+              <span className="text-[11px] text-text-muted shrink-0">{combo.models.length}</span>
+              <span className="material-symbols-outlined text-text-muted text-[16px]">chevron_right</span>
+            </div>
+          </Card>
+        </Link>
+      ))}
+    </div>
+  );
+}
+
+function Section({ title, icon, kind, providers, connections, combos, onCreateCombo }) {
+  return (
+    <div>
+      {/* Header — title left, Create Combo right */}
+      <div className="flex items-center justify-between mb-3">
+        <div className="flex items-center gap-2">
+          <span className="material-symbols-outlined text-primary">{icon}</span>
+          <h2 className="text-base font-semibold">{title}</h2>
+          <span className="text-xs text-text-muted">({providers.length} providers · {combos.length} combos)</span>
+        </div>
+        <Button size="sm" icon="add" onClick={onCreateCombo}>Create Combo</Button>
+      </div>
+
+      {/* Combos — top */}
+      {combos.length > 0 && (
+        <div className="mb-4">
+          <ComboList combos={combos} />
+        </div>
+      )}
+
+      {/* Providers grid — bottom */}
+      {providers.length === 0 ? (
+        <div className="text-center py-8 border border-dashed border-border rounded-xl text-text-muted text-sm">
+          No providers.
+        </div>
+      ) : (
+        <div className="grid grid-cols-1 sm:grid-cols-2 lg:grid-cols-3 xl:grid-cols-4 gap-4">
+          {providers.map((p) => (
+            <ProviderCard key={p.id} provider={p} kind={kind} connections={connections} />
+          ))}
+        </div>
+      )}
+    </div>
+  );
+}
+
+export default function WebProvidersPage() {
+  const router = useRouter();
+  const [connections, setConnections] = useState([]);
+  const [combos, setCombos] = useState([]);
+
+  const fetchAll = async () => {
+    try {
+      const [connsRes, combosRes] = await Promise.all([
+        fetch("/api/providers", { cache: "no-store" }),
+        fetch("/api/combos", { cache: "no-store" }),
+      ]);
+      if (connsRes.ok) setConnections((await connsRes.json()).connections || []);
+      if (combosRes.ok) setCombos((await combosRes.json()).combos || []);
+    } catch { /* noop */ }
+  };
+
+  // eslint-disable-next-line react-hooks/set-state-in-effect
+  useEffect(() => { fetchAll(); }, []);
+
+  const searchProviders = getProvidersByKind("webSearch");
+  const fetchProviders = getProvidersByKind("webFetch");
+  const searchCombos = combos.filter((c) => c.kind === "webSearch");
+  const fetchCombos = combos.filter((c) => c.kind === "webFetch");
+
+  const handleCreateCombo = async (kind) => {
+    // Generate unique default name
+    const base = kind === "webSearch" ? "search-combo" : "fetch-combo";
+    let name = base;
+    let i = 1;
+    const existing = new Set(combos.map((c) => c.name));
+    while (existing.has(name)) { name = `${base}-${i++}`; }
+    const res = await fetch("/api/combos", {
+      method: "POST",
+      headers: { "Content-Type": "application/json" },
+      body: JSON.stringify({ name, models: [], kind }),
+    });
+    if (res.ok) {
+      const created = await res.json();
+      router.push(`/dashboard/media-providers/web/combo/${created.id}`);
+    } else {
+      const err = await res.json();
+      alert(err.error || "Failed to create combo");
+    }
+  };
+
+  return (
+    <div className="flex flex-col gap-8">
+      <Section
+        title="Web Search" icon="search" kind="webSearch"
+        providers={searchProviders} connections={connections} combos={searchCombos}
+        onCreateCombo={() => handleCreateCombo("webSearch")}
+      />
+
+      {/* Divider between sections */}
+      <div className="border-t border-border" />
+
+      <Section
+        title="Web Fetch" icon="travel_explore" kind="webFetch"
+        providers={fetchProviders} connections={connections} combos={fetchCombos}
+        onCreateCombo={() => handleCreateCombo("webFetch")}
+      />
+    </div>
+  );
+}
--- a/src/app/api/combos/route.js
+++ b/src/app/api/combos/route.js
@@ -21,7 +21,7 @@ export async function GET() {
 export async function POST(request) {
  try {
    const body = await request.json();
-    const { name, models } = body;
+    const { name, models, kind } = body;

    if (!name) {
      return NextResponse.json({ error: "Name is required" }, { status: 400 });
@@ -38,7 +38,7 @@ export async function POST(request) {
      return NextResponse.json({ error: "Combo name already exists" }, { status: 400 });
    }

-    const combo = await createCombo({ name, models: models || [] });
+    const combo = await createCombo({ name, models: models || [], kind: kind || null });

    return NextResponse.json(combo, { status: 201 });
  } catch (error) {
--- a/src/app/api/providers/validate/route.js
+++ b/src/app/api/providers/validate/route.js
@@ -1,17 +1,53 @@
 import { NextResponse } from "next/server";
 import { getProviderNodeById } from "@/models";
-import { isOpenAICompatibleProvider, isAnthropicCompatibleProvider, isCustomEmbeddingProvider } from "@/shared/constants/providers";
+import { isOpenAICompatibleProvider, isAnthropicCompatibleProvider, isCustomEmbeddingProvider, AI_PROVIDERS } from "@/shared/constants/providers";
 import { getDefaultModel } from "open-sse/config/providerModels.js";
 import { resolveOllamaLocalHost } from "open-sse/config/providers.js";
 import { PROVIDER_ENDPOINTS } from "@/shared/constants/config";

+// Probe a webSearch/webFetch provider using its searchConfig/fetchConfig.
+// Returns true if API key is accepted (status !== 401 && !== 403).
+async function probeWebProvider(provider, apiKey) {
+  const p = AI_PROVIDERS[provider];
+  if (!p) return null;
+  // Skip if provider has dual-purpose (LLM + search), let LLM validate handle it
+  const kinds = p.serviceKinds || ["llm"];
+  const isWebOnly = kinds.every((k) => k === "webSearch" || k === "webFetch");
+  if (!isWebOnly) return null;
+  const cfg = p.searchConfig || p.fetchConfig;
+  if (!cfg) return null;
+  if (cfg.authType === "none") return true; // no-auth (e.g. searxng)
+
+  let url = cfg.baseUrl;
+  const headers = { "Content-Type": "application/json" };
+  let body;
+
+  // Apply auth based on authHeader
+  switch (cfg.authHeader) {
+    case "bearer":              headers["Authorization"] = `Bearer ${apiKey}`; break;
+    case "x-api-key":           headers["x-api-key"] = apiKey; break;
+    case "x-subscription-token":headers["x-subscription-token"] = apiKey; break;
+    case "key":                 url += `?key=${encodeURIComponent(apiKey)}&q=ping&cx=test`; break; // google-pse
+    case "api_key":             url += `?api_key=${encodeURIComponent(apiKey)}&q=ping&engine=google`; break; // searchapi
+  }
+
+  // Minimal body for POST endpoints; GET sends nothing
+  if (cfg.method === "POST") {
+    body = JSON.stringify({ query: "ping", q: "ping", url: "https://example.com" });
+  }
+
+  const res = await fetch(url, { method: cfg.method, headers, body, signal: AbortSignal.timeout(8000) });
+  return res.status !== 401 && res.status !== 403;
+}
+
 // POST /api/providers/validate - Validate API key with provider
 export async function POST(request) {
  try {
    const body = await request.json();
    const { provider, apiKey, providerSpecificData } = body;

-    if (!provider || (!apiKey && provider !== "ollama-local")) {
+    const isNoAuth = AI_PROVIDERS[provider]?.noAuth === true;
+    if (!provider || (!apiKey && provider !== "ollama-local" && !isNoAuth)) {
      return NextResponse.json({ error: "Provider and API key required" }, { status: 400 });
    }

@@ -147,6 +183,15 @@ export async function POST(request) {
        });
      }

+      // Generic probe for webSearch/webFetch providers (config-driven)
+      const webResult = await probeWebProvider(provider, apiKey);
+      if (webResult !== null) {
+        return NextResponse.json({
+          valid: webResult,
+          error: webResult ? null : "Invalid API key",
+        });
+      }
+
      switch (provider) {
        case "openai":
          const openaiRes = await fetch("https://api.openai.com/v1/models", {
@@ -294,7 +339,12 @@ export async function POST(request) {
          const headers = {};
          if (apiKey) headers["Authorization"] = `Bearer ${apiKey}`;
          const res = await fetch(endpoints[provider], { headers });
-          isValid = res.ok;
+          // xai returns 400 for bad key, 403 for valid-but-no-credit. Other providers use 401.
+          if (provider === "xai") {
+            isValid = res.status === 200 || res.status === 403;
+          } else {
+            isValid = res.ok;
+          }
          break;
        }

--- a/src/app/api/usage/[connectionId]/route.js
+++ b/src/app/api/usage/[connectionId]/route.js
@@ -4,6 +4,7 @@ import "open-sse/index.js";
 import { getProviderConnectionById, updateProviderConnection } from "@/lib/localDb";
 import { getUsageForProvider } from "open-sse/services/usage.js";
 import { getExecutor } from "open-sse/executors/index.js";
+import { resolveConnectionProxyConfig } from "@/lib/network/connectionProxy";

 // Detect auth-expired messages returned by usage providers instead of throwing
 const AUTH_EXPIRED_PATTERNS = ["expired", "authentication", "unauthorized", "401", "re-authorize"];
@@ -18,7 +19,7 @@ function isAuthExpiredMessage(usage) {
 * @param {boolean} force - Skip needsRefresh check and always attempt refresh
 * @returns Promise<{ connection, refreshed: boolean }>
 */
-async function refreshAndUpdateCredentials(connection, force = false) {
+async function refreshAndUpdateCredentials(connection, force = false, proxyOptions = null) {
  const executor = getExecutor(connection.provider);

  // Build credentials object from connection
@@ -39,8 +40,8 @@ async function refreshAndUpdateCredentials(connection, force = false) {
    return { connection, refreshed: false };
  }

-  // Use executor's refreshCredentials method
-  const refreshResult = await executor.refreshCredentials(credentials, console);
+  // Use executor's refreshCredentials method (with optional proxy)
+  const refreshResult = await executor.refreshCredentials(credentials, console, proxyOptions);

  if (!refreshResult) {
    // Refresh failed but we still have an accessToken — try with existing token
@@ -117,9 +118,19 @@ export async function GET(request, { params }) {
      return Response.json({ message: "Usage not available for API key connections" });
    }

+    // Resolve connection proxy config; force strictProxy=false so quota/refresh fall back to direct on failure
+    const proxyConfig = await resolveConnectionProxyConfig(connection.providerSpecificData);
+    const proxyOptions = {
+      connectionProxyEnabled: proxyConfig.connectionProxyEnabled === true,
+      connectionProxyUrl: proxyConfig.connectionProxyUrl || "",
+      connectionNoProxy: proxyConfig.connectionNoProxy || "",
+      vercelRelayUrl: proxyConfig.vercelRelayUrl || "",
+      strictProxy: false,
+    };
+
    // Refresh credentials if needed using executor
    try {
-      const result = await refreshAndUpdateCredentials(connection);
+      const result = await refreshAndUpdateCredentials(connection, false, proxyOptions);
      connection = result.connection;
    } catch (refreshError) {
      console.error("[Usage API] Credential refresh failed:", refreshError);
@@ -129,15 +140,15 @@ export async function GET(request, { params }) {
    }

    // Fetch usage from provider API
-    let usage = await getUsageForProvider(connection);
+    let usage = await getUsageForProvider(connection, proxyOptions);

    // If provider returned an auth-expired message instead of throwing,
    // force-refresh token and retry once
    if (isAuthExpiredMessage(usage) && connection.refreshToken) {
      try {
-        const retryResult = await refreshAndUpdateCredentials(connection, true);
+        const retryResult = await refreshAndUpdateCredentials(connection, true, proxyOptions);
        connection = retryResult.connection;
-        usage = await getUsageForProvider(connection);
+        usage = await getUsageForProvider(connection, proxyOptions);
      } catch (retryError) {
        console.warn(`[Usage] ${connection.provider}: force refresh failed: ${retryError.message}`);
      }
--- a/src/app/api/v1/search/route.js
+++ b/src/app/api/v1/search/route.js
@@ -0,0 +1,21 @@
+import { handleSearch } from "@/sse/handlers/search.js";
+
+/**
+ * Handle CORS preflight
+ */
+export async function OPTIONS() {
+  return new Response(null, {
+    headers: {
+      "Access-Control-Allow-Origin": "*",
+      "Access-Control-Allow-Methods": "POST, OPTIONS",
+      "Access-Control-Allow-Headers": "*"
+    }
+  });
+}
+
+/**
+ * POST /v1/search - Web search endpoint
+ */
+export async function POST(request) {
+  return await handleSearch(request);
+}
--- a/src/app/api/v1/web/fetch/route.js
+++ b/src/app/api/v1/web/fetch/route.js
@@ -0,0 +1,21 @@
+import { handleFetch } from "@/sse/handlers/fetch.js";
+
+/**
+ * Handle CORS preflight
+ */
+export async function OPTIONS() {
+  return new Response(null, {
+    headers: {
+      "Access-Control-Allow-Origin": "*",
+      "Access-Control-Allow-Methods": "POST, OPTIONS",
+      "Access-Control-Allow-Headers": "*"
+    }
+  });
+}
+
+/**
+ * POST /v1/web/fetch - Web URL fetch/extract endpoint
+ */
+export async function POST(request) {
+  return await handleFetch(request);
+}
--- a/src/lib/localDb.js
+++ b/src/lib/localDb.js
@@ -582,6 +582,7 @@ export async function createCombo(data) {
    id: uuidv4(),
    name: data.name,
    models: data.models || [],
+    kind: data.kind || null,
    createdAt: now,
    updatedAt: now,
  };
--- a/src/shared/components/ModelSelectModal.js
+++ b/src/shared/components/ModelSelectModal.js
@@ -4,7 +4,7 @@ import { useState, useMemo, useEffect } from "react";
 import PropTypes from "prop-types";
 import Modal from "./Modal";
 import { getModelsByProviderId } from "@/shared/constants/models";
-import { OAUTH_PROVIDERS, APIKEY_PROVIDERS, FREE_PROVIDERS, FREE_TIER_PROVIDERS, isOpenAICompatibleProvider, isAnthropicCompatibleProvider, getProviderAlias } from "@/shared/constants/providers";
+import { OAUTH_PROVIDERS, APIKEY_PROVIDERS, FREE_PROVIDERS, FREE_TIER_PROVIDERS, AI_PROVIDERS, isOpenAICompatibleProvider, isAnthropicCompatibleProvider, getProviderAlias } from "@/shared/constants/providers";

 // Provider order: OAuth first, then Free Tier, then API Key (matches dashboard/providers)
 const PROVIDER_ORDER = [
@@ -25,7 +25,17 @@ export default function ModelSelectModal({
  activeProviders = [],
  title = "Select Model",
  modelAliases = {},
+  kindFilter = null,
 }) {
+  // Filter activeProviders by serviceKinds when kindFilter set (e.g. "webSearch", "webFetch")
+  const filteredActiveProviders = useMemo(() => {
+    if (!kindFilter) return activeProviders;
+    return activeProviders.filter((p) => {
+      const info = AI_PROVIDERS[p.provider];
+      const kinds = info?.serviceKinds || ["llm"];
+      return kinds.includes(kindFilter);
+    });
+  }, [activeProviders, kindFilter]);
  const [searchQuery, setSearchQuery] = useState("");
  const [combos, setCombos] = useState([]);
  const [providerNodes, setProviderNodes] = useState([]);
@@ -85,13 +95,18 @@ export default function ModelSelectModal({
  const groupedModels = useMemo(() => {
    const groups = {};

-    // Get all active provider IDs from connections
-    const activeConnectionIds = activeProviders.map(p => p.provider);
+    // Get all active provider IDs from connections (filtered by kindFilter if set)
+    const activeConnectionIds = filteredActiveProviders.map(p => p.provider);
+
+    // No-auth providers: filter by kindFilter as well
+    const noAuthIds = kindFilter
+      ? NO_AUTH_PROVIDER_IDS.filter((id) => (AI_PROVIDERS[id]?.serviceKinds || ["llm"]).includes(kindFilter))
+      : NO_AUTH_PROVIDER_IDS;

    // Only show connected providers (including both standard and custom)
    const providerIdsToShow = new Set([
      ...activeConnectionIds,  // Only connected providers
-      ...NO_AUTH_PROVIDER_IDS, // No-auth providers always visible
+      ...noAuthIds,            // No-auth providers (kind-filtered)
    ]);

    // Sort by PROVIDER_ORDER
@@ -203,14 +218,15 @@ export default function ModelSelectModal({
    });

    return groups;
-  }, [activeProviders, modelAliases, allProviders, providerNodes, customModels]);
+  }, [filteredActiveProviders, modelAliases, allProviders, providerNodes, customModels, kindFilter]);

-  // Filter combos by search query
+  // Filter combos by search query (and hide combos when kindFilter is set — combos are LLM-only by design)
  const filteredCombos = useMemo(() => {
+    if (kindFilter) return [];
    if (!searchQuery.trim()) return combos;
    const query = searchQuery.toLowerCase();
    return combos.filter(c => c.name.toLowerCase().includes(query));
-  }, [combos, searchQuery]);
+  }, [combos, searchQuery, kindFilter]);

  // Filter models by search query
  const filteredGroups = useMemo(() => {
@@ -384,5 +400,6 @@ ModelSelectModal.propTypes = {
  ),
  title: PropTypes.string,
  modelAliases: PropTypes.object,
+  kindFilter: PropTypes.string,
 };

--- a/src/shared/components/ProviderInfoCard.js
+++ b/src/shared/components/ProviderInfoCard.js
@@ -2,24 +2,20 @@

 import Card from "./Card";

-// Field schema — config-driven, used for both searchConfig and fetchConfig
+// Only show fields user actually cares about
 const FIELD_SCHEMA = {
+  mode:             { label: "Mode",       format: (v) => v },
+  defaultModel:     { label: "Model",      format: (v) => v, mono: true },
  baseUrl:          { label: "Endpoint",   format: (v) => v, isLink: true, mono: true },
-  method:           { label: "Method",     format: (v) => v },
-  authType:         { label: "Auth",       format: (v) => v },
-  authHeader:       { label: "Auth Header", format: (v) => v, mono: true },
  costPerQuery:     { label: "Cost / call", format: (v) => v === 0 ? "Free" : `$${v.toFixed(4)}` },
  freeMonthlyQuota: { label: "Free quota",  format: (v) => v === 0 ? "—" : v >= 999999 ? "Unlimited" : `${v.toLocaleString()} / mo` },
  searchTypes:      { label: "Types",      format: (v) => v.join(", ") },
  formats:          { label: "Formats",    format: (v) => v.join(", ") },
-  defaultMaxResults: { label: "Default results", format: (v) => v },
  maxMaxResults:    { label: "Max results", format: (v) => v },
  maxCharacters:    { label: "Max chars",  format: (v) => v.toLocaleString() },
-  timeoutMs:        { label: "Timeout",    format: (v) => `${v / 1000}s` },
-  cacheTTLMs:       { label: "Cache TTL",  format: (v) => `${v / 60000}m` },
 };

-export default function ProviderInfoCard({ config, title = "Provider Info" }) {
+export default function ProviderInfoCard({ config, provider, title = "Provider Info" }) {
  if (!config) return null;

  const rows = Object.entries(FIELD_SCHEMA)
@@ -33,9 +29,24 @@ export default function ProviderInfoCard({ config, title = "Provider Info" }) {
      raw: config[key],
    }));

+  const signupUrl = provider?.notice?.apiKeyUrl || provider?.website;
+
  return (
    <Card>
-      <h2 className="text-lg font-semibold mb-3">{title}</h2>
+      <div className="flex items-center justify-between mb-3">
+        <h2 className="text-lg font-semibold">{title}</h2>
+        {signupUrl && (
+          <a
+            href={signupUrl}
+            target="_blank"
+            rel="noopener noreferrer"
+            className="text-xs text-primary hover:underline inline-flex items-center gap-1"
+          >
+            <span className="material-symbols-outlined text-sm">open_in_new</span>
+            Get API Key
+          </a>
+        )}
+      </div>
      <div className="grid grid-cols-1 sm:grid-cols-2 gap-x-6 gap-y-2">
        {rows.map((r) => (
          <div key={r.key} className="flex items-center gap-3 min-w-0">
--- a/src/shared/components/Sidebar.js
+++ b/src/shared/components/Sidebar.js
@@ -12,7 +12,9 @@ import Button from "./Button";
 import { ConfirmModal } from "./Modal";

 // const VISIBLE_MEDIA_KINDS = ["embedding", "image", "imageToText", "tts", "stt", "webSearch", "webFetch", "video", "music"];
-const VISIBLE_MEDIA_KINDS = ["embedding", "image", "tts", "webSearch", "webFetch"];
+const VISIBLE_MEDIA_KINDS = ["embedding", "image", "tts"];
+// Combined entry: webSearch + webFetch share one page at /dashboard/media-providers/web
+const COMBINED_WEB_ITEM = { id: "web", label: "Web Fetch & Search", icon: "travel_explore", href: "/dashboard/media-providers/web" };

 const navItems = [
  { href: "/dashboard/endpoint", label: "Endpoint", icon: "api" },
@@ -234,6 +236,20 @@ export default function Sidebar({ onClose }) {
                    <span className="text-sm">{kind.label}</span>
                  </Link>
                ))}
+                <Link
+                  key={COMBINED_WEB_ITEM.id}
+                  href={COMBINED_WEB_ITEM.href}
+                  onClick={onClose}
+                  className={cn(
+                    "flex items-center gap-3 px-4 py-1.5 rounded-lg transition-all group",
+                    pathname.startsWith(COMBINED_WEB_ITEM.href)
+                      ? "bg-primary/10 text-primary"
+                      : "text-text-muted hover:bg-surface/50 hover:text-text-main"
+                  )}
+                >
+                  <span className="material-symbols-outlined text-[16px]">{COMBINED_WEB_ITEM.icon}</span>
+                  <span className="text-sm">{COMBINED_WEB_ITEM.label}</span>
+                </Link>
              </div>
            )}

--- a/src/shared/constants/providers.js
+++ b/src/shared/constants/providers.js
@@ -18,7 +18,7 @@ export const FREE_TIER_PROVIDERS = {
  nvidia: { id: "nvidia", alias: "nvidia", name: "NVIDIA NIM", icon: "developer_board", color: "#76B900", textIcon: "NV", website: "https://developer.nvidia.com/nim", notice: { text: "Free access for NVIDIA Developer Program members (prototyping & testing).", apiKeyUrl: "https://build.nvidia.com/settings/api-keys" } },
  ollama: { id: "ollama", alias: "ollama", name: "Ollama Cloud", icon: "cloud", color: "#ffffffff", textIcon: "OL", website: "https://ollama.com", notice: { text: "Free tier: light usage, 1 cloud model at a time (limits reset every 5h & 7d). Pro $20/mo · Max $100/mo.", apiKeyUrl: "https://ollama.com/settings/keys" } },
  vertex: { id: "vertex", alias: "vx", name: "Vertex AI", icon: "cloud", color: "#4285F4", textIcon: "VX", website: "https://cloud.google.com/vertex-ai", notice: { text: "New Google Cloud accounts get $300 free credits. Requires GCP project + Service Account with Vertex AI API enabled.", apiKeyUrl: "https://console.cloud.google.com/iam-admin/serviceaccounts" } },
-  gemini: { id: "gemini", alias: "gemini", name: "Gemini", icon: "diamond", color: "#4285F4", textIcon: "GE", website: "https://ai.google.dev", serviceKinds: ["llm", "embedding", "image", "imageToText", "webSearch"], searchConfig: { baseUrl: "https://generativelanguage.googleapis.com", method: "POST", authType: "apikey", authHeader: "x-goog-api-key", costPerQuery: 0, freeMonthlyQuota: 1500, searchTypes: ["web"], defaultMaxResults: 5, maxMaxResults: 10, timeoutMs: 15000, cacheTTLMs: 300000 } },
+  gemini: { id: "gemini", alias: "gemini", name: "Gemini", icon: "diamond", color: "#4285F4", textIcon: "GE", website: "https://ai.google.dev", serviceKinds: ["llm", "embedding", "image", "imageToText", "webSearch"], searchViaChat: { defaultModel: "gemini-2.5-flash" } },
  byteplus: { id: "byteplus", alias: "bpm", name: "BytePlus ModelArk", icon: "cloud", color: "#2563EB", textIcon: "BP", website: "https://console.byteplus.com/ark", notice: { text: "Free credits for new accounts. Access to Seed 2.0, Kimi K2 Thinking, GLM 4.7, GPT-OSS-120B models.", apiKeyUrl: "https://console.byteplus.com/ark/region:ark+ap-southeast-1/apiKey" }, serviceKinds: ["llm"] },
 };

@@ -55,20 +55,20 @@ export const OAUTH_PROVIDERS = {
 export const APIKEY_PROVIDERS = {
  glm: { id: "glm", alias: "glm", name: "GLM Coding", icon: "code", color: "#2563EB", textIcon: "GL", website: "https://open.bigmodel.cn" },
  "glm-cn": { id: "glm-cn", alias: "glm-cn", name: "GLM (China)", icon: "code", color: "#DC2626", textIcon: "GC", website: "https://open.bigmodel.cn" },
-  kimi: { id: "kimi", alias: "kimi", name: "Kimi", icon: "psychology", color: "#1E3A8A", textIcon: "KM", website: "https://kimi.moonshot.cn", serviceKinds: ["llm", "webSearch"], searchConfig: { baseUrl: "https://api.moonshot.cn", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0, freeMonthlyQuota: 0, searchTypes: ["web"], defaultMaxResults: 5, maxMaxResults: 10, timeoutMs: 15000, cacheTTLMs: 300000 } },
-  minimax: { id: "minimax", alias: "minimax", name: "Minimax Coding", icon: "memory", color: "#7C3AED", textIcon: "MM", website: "https://www.minimaxi.com", serviceKinds: ["llm", "image", "imageToText", "webSearch"], searchConfig: { baseUrl: "https://api.minimaxi.com", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0, freeMonthlyQuota: 0, searchTypes: ["web"], defaultMaxResults: 5, maxMaxResults: 10, timeoutMs: 15000, cacheTTLMs: 300000 } },
+  kimi: { id: "kimi", alias: "kimi", name: "Kimi", icon: "psychology", color: "#1E3A8A", textIcon: "KM", website: "https://kimi.moonshot.cn", serviceKinds: ["llm", "webSearch"], searchViaChat: { defaultModel: "kimi-k2.5" } },
+  minimax: { id: "minimax", alias: "minimax", name: "Minimax Coding", icon: "memory", color: "#7C3AED", textIcon: "MM", website: "https://www.minimaxi.com", serviceKinds: ["llm", "image", "imageToText", "webSearch"], searchViaChat: { defaultModel: "MiniMax-M2.7" } },
  "minimax-cn": { id: "minimax-cn", alias: "minimax-cn", name: "Minimax (China)", icon: "memory", color: "#DC2626", textIcon: "MC", website: "https://www.minimaxi.com" },
  alicode: { id: "alicode", alias: "alicode", name: "Alibaba", icon: "cloud", color: "#FF6A00", textIcon: "ALi" },
  "alicode-intl": { id: "alicode-intl", alias: "alicode-intl", name: "Alibaba Intl", icon: "cloud", color: "#FF6A00", textIcon: "ALi" },
  "volcengine-ark": { id: "volcengine-ark", alias: "ark", name: "Volcengine Ark", icon: "cloud", color: "#1677FF", textIcon: "ARK", website: "https://ark.cn-beijing.volces.com" },
-  openai: { id: "openai", alias: "openai", name: "OpenAI", icon: "auto_awesome", color: "#10A37F", textIcon: "OA", website: "https://platform.openai.com", serviceKinds: ["llm", "embedding", "tts", "image", "imageToText", "webSearch"], thinkingConfig: THINKING_CONFIG.effort, searchConfig: { baseUrl: "https://api.openai.com", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.025, freeMonthlyQuota: 0, searchTypes: ["web"], defaultMaxResults: 5, maxMaxResults: 10, timeoutMs: 15000, cacheTTLMs: 300000 } },
+  openai: { id: "openai", alias: "openai", name: "OpenAI", icon: "auto_awesome", color: "#10A37F", textIcon: "OA", website: "https://platform.openai.com", serviceKinds: ["llm", "embedding", "tts", "image", "imageToText", "webSearch"], thinkingConfig: THINKING_CONFIG.effort, searchViaChat: { defaultModel: "gpt-4o-mini" } },
  anthropic: { id: "anthropic", alias: "anthropic", name: "Anthropic", icon: "smart_toy", color: "#D97757", textIcon: "AN", website: "https://console.anthropic.com", serviceKinds: ["llm", "imageToText"] },
  "opencode-go": { id: "opencode-go", alias: "ocg", name: "OpenCode Go", icon: "terminal", color: "#E87040", textIcon: "OC", website: "https://opencode.ai/auth", notice: { text: "OpenCode Go subscription: $5/mo (then $10/mo). Access to Kimi, GLM, Qwen, MiMo, MiniMax models.", apiKeyUrl: "https://opencode.ai/auth" } },
  azure: { id: "azure", alias: "azure", name: "Azure OpenAI", icon: "cloud", color: "#0078D4", textIcon: "AZ", website: "https://azure.microsoft.com/en-us/products/ai-services/openai-service", hasProviderSpecificData: true },

  deepseek: { id: "deepseek", alias: "ds", name: "DeepSeek", icon: "bolt", color: "#4D6BFE", textIcon: "DS", website: "https://deepseek.com" },
  groq: { id: "groq", alias: "groq", name: "Groq", icon: "speed", color: "#F55036", textIcon: "GQ", website: "https://groq.com", serviceKinds: ["llm", "imageToText"] },
-  xai: { id: "xai", alias: "xai", name: "xAI (Grok)", icon: "auto_awesome", color: "#1DA1F2", textIcon: "XA", website: "https://x.ai", serviceKinds: ["llm", "imageToText", "webSearch"], searchConfig: { baseUrl: "https://api.x.ai", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.025, freeMonthlyQuota: 0, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 30, timeoutMs: 15000, cacheTTLMs: 300000 } },
+  xai: { id: "xai", alias: "xai", name: "xAI (Grok)", icon: "auto_awesome", color: "#1DA1F2", textIcon: "XA", website: "https://x.ai", serviceKinds: ["llm", "imageToText", "webSearch"], searchViaChat: { defaultModel: "grok-4.20-reasoning" } },
  mistral: { id: "mistral", alias: "mistral", name: "Mistral", icon: "air", color: "#FF7000", textIcon: "MI", website: "https://mistral.ai", serviceKinds: ["llm", "imageToText"] },
  perplexity: { id: "perplexity", alias: "pplx", name: "Perplexity", icon: "search", color: "#20808D", textIcon: "PP", website: "https://www.perplexity.ai", serviceKinds: ["llm", "webSearch"], searchConfig: { baseUrl: "https://api.perplexity.ai/search", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.005, freeMonthlyQuota: 0, searchTypes: ["web"], defaultMaxResults: 5, maxMaxResults: 20, timeoutMs: 10000, cacheTTLMs: 300000 } },
  together: { id: "together", alias: "together", name: "Together AI", icon: "group_work", color: "#0F6FFF", textIcon: "TG", website: "https://www.together.ai" },
@@ -94,17 +94,17 @@ export const APIKEY_PROVIDERS = {
  chutes: { id: "chutes", alias: "ch", name: "Chutes AI", icon: "water_drop", color: "#ffffffff", textIcon: "CH", website: "https://chutes.ai" },
  "ollama-local": { id: "ollama-local", alias: "ollama-local", name: "Ollama Local", icon: "cloud", color: "#ffffffff", textIcon: "OL", website: "https://ollama.com" },
  "vertex-partner": { id: "vertex-partner", alias: "vxp", name: "Vertex Partner", icon: "cloud", color: "#34A853", textIcon: "VP", website: "https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-partner-models" },
-  tavily: { id: "tavily", alias: "tavily", name: "Tavily", icon: "search", color: "#5B21B6", textIcon: "TV", website: "https://tavily.com", serviceKinds: ["webSearch", "webFetch"], searchConfig: { baseUrl: "https://api.tavily.com/search", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.008, freeMonthlyQuota: 1000, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 20, timeoutMs: 10000, cacheTTLMs: 300000 }, fetchConfig: { baseUrl: "https://api.tavily.com/extract", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.008, freeMonthlyQuota: 1000, formats: ["markdown", "text"], maxCharacters: 100000, timeoutMs: 15000 } },
-  "brave-search": { id: "brave-search", alias: "brave", name: "Brave Search", icon: "travel_explore", color: "#FB542B", textIcon: "BR", website: "https://brave.com/search/api", serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://api.search.brave.com/res/v1", method: "GET", authType: "apikey", authHeader: "x-subscription-token", costPerQuery: 0.005, freeMonthlyQuota: 1000, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 20, timeoutMs: 10000, cacheTTLMs: 300000 } },
-  serper: { id: "serper", alias: "serper", name: "Serper", icon: "search", color: "#4F46E5", textIcon: "SP", website: "https://serper.dev", serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://google.serper.dev", method: "POST", authType: "apikey", authHeader: "x-api-key", costPerQuery: 0.001, freeMonthlyQuota: 2500, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 100, timeoutMs: 10000, cacheTTLMs: 300000 } },
-  exa: { id: "exa", alias: "exa", name: "Exa", icon: "manage_search", color: "#2563EB", textIcon: "EX", website: "https://exa.ai", serviceKinds: ["webSearch", "webFetch"], searchConfig: { baseUrl: "https://api.exa.ai/search", method: "POST", authType: "apikey", authHeader: "x-api-key", costPerQuery: 0.007, freeMonthlyQuota: 1000, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 100, timeoutMs: 10000, cacheTTLMs: 300000 }, fetchConfig: { baseUrl: "https://api.exa.ai/contents", method: "POST", authType: "apikey", authHeader: "x-api-key", costPerQuery: 0.001, freeMonthlyQuota: 1000, formats: ["text", "markdown"], maxCharacters: 100000, timeoutMs: 15000 } },
+  tavily: { id: "tavily", alias: "tavily", name: "Tavily", icon: "search", color: "#5B21B6", textIcon: "TV", website: "https://tavily.com", notice: { apiKeyUrl: "https://app.tavily.com/home" }, serviceKinds: ["webSearch", "webFetch"], searchConfig: { baseUrl: "https://api.tavily.com/search", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.008, freeMonthlyQuota: 1000, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 20, timeoutMs: 10000, cacheTTLMs: 300000 }, fetchConfig: { baseUrl: "https://api.tavily.com/extract", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.008, freeMonthlyQuota: 1000, formats: ["markdown", "text"], maxCharacters: 100000, timeoutMs: 15000 } },
+  "brave-search": { id: "brave-search", alias: "brave", name: "Brave Search", icon: "travel_explore", color: "#FB542B", textIcon: "BR", website: "https://brave.com/search/api", notice: { apiKeyUrl: "https://api-dashboard.search.brave.com/app/keys" }, serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://api.search.brave.com/res/v1", method: "GET", authType: "apikey", authHeader: "x-subscription-token", costPerQuery: 0.005, freeMonthlyQuota: 1000, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 20, timeoutMs: 10000, cacheTTLMs: 300000 } },
+  serper: { id: "serper", alias: "serper", name: "Serper", icon: "search", color: "#4F46E5", textIcon: "SP", website: "https://serper.dev", notice: { apiKeyUrl: "https://serper.dev/api-key" }, serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://google.serper.dev", method: "POST", authType: "apikey", authHeader: "x-api-key", costPerQuery: 0.001, freeMonthlyQuota: 2500, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 100, timeoutMs: 10000, cacheTTLMs: 300000 } },
+  exa: { id: "exa", alias: "exa", name: "Exa", icon: "manage_search", color: "#2563EB", textIcon: "EX", website: "https://exa.ai", notice: { apiKeyUrl: "https://dashboard.exa.ai/api-keys" }, serviceKinds: ["webSearch", "webFetch"], searchConfig: { baseUrl: "https://api.exa.ai/search", method: "POST", authType: "apikey", authHeader: "x-api-key", costPerQuery: 0.007, freeMonthlyQuota: 1000, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 100, timeoutMs: 10000, cacheTTLMs: 300000 }, fetchConfig: { baseUrl: "https://api.exa.ai/contents", method: "POST", authType: "apikey", authHeader: "x-api-key", costPerQuery: 0.001, freeMonthlyQuota: 1000, formats: ["text", "markdown"], maxCharacters: 100000, timeoutMs: 15000 } },
  searxng: { id: "searxng", alias: "searxng", name: "SearXNG", icon: "saved_search", color: "#3B82F6", textIcon: "SX", website: "https://docs.searxng.org", serviceKinds: ["webSearch"], noAuth: true, searchConfig: { baseUrl: "http://localhost:8888/search", method: "GET", authType: "none", authHeader: "none", costPerQuery: 0, freeMonthlyQuota: 999999, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 50, timeoutMs: 10000, cacheTTLMs: 180000 } },
-  "google-pse": { id: "google-pse", alias: "gpse", name: "Google PSE", icon: "search", color: "#4285F4", textIcon: "GP", website: "https://programmablesearchengine.google.com", serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://www.googleapis.com/customsearch/v1", method: "GET", authType: "apikey", authHeader: "key", costPerQuery: 0.005, freeMonthlyQuota: 3000, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 10, timeoutMs: 10000, cacheTTLMs: 300000 } },
-  linkup: { id: "linkup", alias: "linkup", name: "Linkup", icon: "link", color: "#0EA5E9", textIcon: "LK", website: "https://linkup.so", serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://api.linkup.so/v1/search", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.005, freeMonthlyQuota: 1000, searchTypes: ["web"], defaultMaxResults: 5, maxMaxResults: 50, timeoutMs: 10000, cacheTTLMs: 300000 } },
-  searchapi: { id: "searchapi", alias: "searchapi", name: "SearchAPI", icon: "search", color: "#0EA5A4", textIcon: "SA", website: "https://www.searchapi.io", serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://www.searchapi.io/api/v1/search", method: "GET", authType: "apikey", authHeader: "api_key", costPerQuery: 0.004, freeMonthlyQuota: 100, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 100, timeoutMs: 10000, cacheTTLMs: 300000 } },
-  youcom: { id: "youcom", alias: "youcom", name: "You.com Search", icon: "search", color: "#7C3AED", textIcon: "YC", website: "https://you.com", serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://ydc-index.io/v1/search", method: "GET", authType: "apikey", authHeader: "x-api-key", costPerQuery: 0.005, freeMonthlyQuota: 0, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 100, timeoutMs: 10000, cacheTTLMs: 300000 } },
-  firecrawl: { id: "firecrawl", alias: "firecrawl", name: "Firecrawl", icon: "local_fire_department", color: "#F59E0B", textIcon: "FC", website: "https://firecrawl.dev", serviceKinds: ["webFetch"], fetchConfig: { baseUrl: "https://api.firecrawl.dev/v1/scrape", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.002, freeMonthlyQuota: 500, formats: ["markdown", "html", "text"], maxCharacters: 200000, timeoutMs: 30000 } },
-  "jina-reader": { id: "jina-reader", alias: "jina", name: "Jina Reader", icon: "menu_book", color: "#000000", textIcon: "JR", website: "https://jina.ai/reader", serviceKinds: ["webFetch"], fetchConfig: { baseUrl: "https://r.jina.ai", method: "GET", authType: "apikey", authHeader: "bearer", costPerQuery: 0, freeMonthlyQuota: 1000000, formats: ["markdown", "text", "html"], maxCharacters: 200000, timeoutMs: 30000 } },
+  "google-pse": { id: "google-pse", alias: "gpse", name: "Google PSE", icon: "search", color: "#4285F4", textIcon: "GP", website: "https://programmablesearchengine.google.com", notice: { apiKeyUrl: "https://programmablesearchengine.google.com/controlpanel/create" }, serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://www.googleapis.com/customsearch/v1", method: "GET", authType: "apikey", authHeader: "key", costPerQuery: 0.005, freeMonthlyQuota: 3000, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 10, timeoutMs: 10000, cacheTTLMs: 300000 } },
+  linkup: { id: "linkup", alias: "linkup", name: "Linkup", icon: "link", color: "#0EA5E9", textIcon: "LK", website: "https://linkup.so", notice: { apiKeyUrl: "https://app.linkup.so/api-keys" }, serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://api.linkup.so/v1/search", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.005, freeMonthlyQuota: 1000, searchTypes: ["web"], defaultMaxResults: 5, maxMaxResults: 50, timeoutMs: 10000, cacheTTLMs: 300000 } },
+  searchapi: { id: "searchapi", alias: "searchapi", name: "SearchAPI", icon: "search", color: "#0EA5A4", textIcon: "SA", website: "https://www.searchapi.io", notice: { apiKeyUrl: "https://www.searchapi.io/dashboard" }, serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://www.searchapi.io/api/v1/search", method: "GET", authType: "apikey", authHeader: "api_key", costPerQuery: 0.004, freeMonthlyQuota: 100, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 100, timeoutMs: 10000, cacheTTLMs: 300000 } },
+  youcom: { id: "youcom", alias: "youcom", name: "You.com Search", icon: "search", color: "#7C3AED", textIcon: "YC", website: "https://you.com", notice: { apiKeyUrl: "https://api.you.com" }, serviceKinds: ["webSearch"], searchConfig: { baseUrl: "https://ydc-index.io/v1/search", method: "GET", authType: "apikey", authHeader: "x-api-key", costPerQuery: 0.005, freeMonthlyQuota: 0, searchTypes: ["web", "news"], defaultMaxResults: 5, maxMaxResults: 100, timeoutMs: 10000, cacheTTLMs: 300000 } },
+  firecrawl: { id: "firecrawl", alias: "firecrawl", name: "Firecrawl", icon: "local_fire_department", color: "#F59E0B", textIcon: "FC", website: "https://firecrawl.dev", notice: { apiKeyUrl: "https://www.firecrawl.dev/app/api-keys" }, serviceKinds: ["webFetch"], fetchConfig: { baseUrl: "https://api.firecrawl.dev/v1/scrape", method: "POST", authType: "apikey", authHeader: "bearer", costPerQuery: 0.002, freeMonthlyQuota: 500, formats: ["markdown", "html", "text"], maxCharacters: 200000, timeoutMs: 30000 } },
+  "jina-reader": { id: "jina-reader", alias: "jina", name: "Jina Reader", icon: "menu_book", color: "#000000", textIcon: "JR", website: "https://jina.ai/reader", notice: { apiKeyUrl: "https://jina.ai/?sui=apikey" }, serviceKinds: ["webFetch"], fetchConfig: { baseUrl: "https://r.jina.ai", method: "GET", authType: "apikey", authHeader: "bearer", costPerQuery: 0, freeMonthlyQuota: 1000000, formats: ["markdown", "text", "html"], maxCharacters: 200000, timeoutMs: 30000 } },
 };

 // Web Cookie Providers (use browser session cookie instead of API key)
--- a/src/sse/handlers/fetch.js
+++ b/src/sse/handlers/fetch.js
@@ -0,0 +1,211 @@
+import {
+  getProviderCredentials,
+  markAccountUnavailable,
+  clearAccountError,
+  extractApiKey,
+  isValidApiKey,
+} from "../services/auth.js";
+import { getSettings, getCombos } from "@/lib/localDb";
+import { AI_PROVIDERS, resolveProviderId } from "@/shared/constants/providers.js";
+import { handleFetchCore } from "open-sse/handlers/fetch/index.js";
+import { errorResponse, unavailableResponse } from "open-sse/utils/error.js";
+import { HTTP_STATUS } from "open-sse/config/runtimeConfig.js";
+import * as log from "../utils/logger.js";
+import { updateProviderCredentials, checkAndRefreshToken } from "../services/tokenRefresh.js";
+import { handleComboChat, getComboModelsFromData } from "open-sse/services/combo.js";
+
+/**
+ * Handle web fetch (URL extraction) request for the SSE/Next.js server.
+ * Provider IS the model. Mirrors handleEmbeddings auth + fallback flow.
+ *
+ * @param {Request} request
+ */
+export async function handleFetch(request) {
+  let body;
+  try {
+    body = await request.json();
+  } catch {
+    log.warn("FETCH", "Invalid JSON body");
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, "Invalid JSON body");
+  }
+
+  const reqUrl = new URL(request.url);
+  // Accept either `provider` or `model` (UI sends `model` since provider IS the model for webFetch)
+  const providerInput = body.provider || body.model;
+  const targetUrl = body.url;
+  const format = body.format;
+  const maxCharacters = body.max_characters;
+
+  log.request("POST", `${reqUrl.pathname} | ${providerInput}`);
+
+  // Log API key (masked)
+  const apiKey = extractApiKey(request);
+  if (apiKey) {
+    log.debug("AUTH", `API Key: ${log.maskKey(apiKey)}`);
+  } else {
+    log.debug("AUTH", "No API key provided (local mode)");
+  }
+
+  // Enforce API key if enabled in settings
+  const settings = await getSettings();
+  if (settings.requireApiKey) {
+    if (!apiKey) {
+      log.warn("AUTH", "Missing API key (requireApiKey=true)");
+      return errorResponse(HTTP_STATUS.UNAUTHORIZED, "Missing API key");
+    }
+    const valid = await isValidApiKey(apiKey);
+    if (!valid) {
+      log.warn("AUTH", "Invalid API key (requireApiKey=true)");
+      return errorResponse(HTTP_STATUS.UNAUTHORIZED, "Invalid API key");
+    }
+  }
+
+  if (!providerInput || typeof providerInput !== "string") {
+    log.warn("FETCH", "Missing provider/model");
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, "Missing required field: provider (or model)");
+  }
+
+  if (!targetUrl || typeof targetUrl !== "string") {
+    log.warn("FETCH", "Missing url");
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, "Missing required field: url");
+  }
+
+  // Validate URL format
+  try {
+    new URL(targetUrl);
+  } catch {
+    log.warn("FETCH", "Invalid URL", { url: targetUrl });
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, "Invalid URL format");
+  }
+
+  // Combo expansion: providerInput may be a combo name → run fallback/round-robin across providers
+  const combos = await getCombos();
+  const comboModels = getComboModelsFromData(providerInput, combos);
+  if (comboModels) {
+    const comboStrategies = settings.comboStrategies || {};
+    const comboStrategy = comboStrategies[providerInput]?.fallbackStrategy || settings.comboStrategy || "fallback";
+    log.info("FETCH", `Combo "${providerInput}" with ${comboModels.length} providers (strategy: ${comboStrategy})`);
+    return handleComboChat({
+      body,
+      models: comboModels,
+      handleSingleModel: (b, m) => handleSingleProviderFetch(b, m, request, apiKey, settings),
+      log,
+      comboName: providerInput,
+      comboStrategy
+    });
+  }
+
+  return handleSingleProviderFetch(body, providerInput, request, apiKey, settings);
+}
+
+async function handleSingleProviderFetch(body, providerInput, request, apiKey, settings) {
+  const targetUrl = body.url;
+  const format = body.format;
+  const maxCharacters = body.max_characters;
+  const providerId = resolveProviderId(providerInput);
+  const resolvedProvider = AI_PROVIDERS[providerId];
+
+  if (!resolvedProvider) {
+    log.warn("FETCH", "Unknown provider", { provider: providerInput });
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, `Unknown provider: ${providerInput}`);
+  }
+
+  const providerConfig = resolvedProvider.fetchConfig;
+  if (!providerConfig) {
+    log.warn("FETCH", "Provider does not support web fetch", { provider: providerId });
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, `Provider ${providerId} does not support web fetch`);
+  }
+
+  if (providerInput !== providerId) {
+    log.info("ROUTING", `${providerInput} → ${providerId}`);
+  } else {
+    log.info("ROUTING", `Provider: ${providerId}`);
+  }
+
+  // No-auth fetch path (kept for parity though no current fetch provider sets noAuth)
+  if (resolvedProvider.noAuth) {
+    log.info("AUTH", `\x1b[32m${providerId} no-auth mode\x1b[0m`);
+    const result = await handleFetchCore({
+      url: targetUrl,
+      format,
+      maxCharacters,
+      provider: resolvedProvider.id,
+      providerConfig,
+      credentials: null,
+      log
+    });
+    if (result.success) {
+      return new Response(JSON.stringify(result.data), {
+        headers: { "Content-Type": "application/json", "Access-Control-Allow-Origin": "*" }
+      });
+    }
+    return errorResponse(result.status || HTTP_STATUS.BAD_GATEWAY, result.error || "Fetch failed");
+  }
+
+  // Credential + fallback loop
+  const excludeConnectionIds = new Set();
+  let lastError = null;
+  let lastStatus = null;
+
+  while (true) {
+    const credentials = await getProviderCredentials(providerId, excludeConnectionIds);
+
+    if (!credentials || credentials.allRateLimited) {
+      if (credentials?.allRateLimited) {
+        const errorMsg = lastError || credentials.lastError || "Unavailable";
+        const status = lastStatus || Number(credentials.lastErrorCode) || HTTP_STATUS.SERVICE_UNAVAILABLE;
+        log.warn("FETCH", `[${providerId}] ${errorMsg} (${credentials.retryAfterHuman})`);
+        return unavailableResponse(status, `[${providerId}] ${errorMsg}`, credentials.retryAfter, credentials.retryAfterHuman);
+      }
+      if (excludeConnectionIds.size === 0) {
+        log.error("AUTH", `No credentials for provider: ${providerId}`);
+        return errorResponse(HTTP_STATUS.BAD_REQUEST, `No credentials for provider: ${providerId}`);
+      }
+      log.warn("FETCH", "No more accounts available", { provider: providerId });
+      return errorResponse(lastStatus || HTTP_STATUS.SERVICE_UNAVAILABLE, lastError || "All accounts unavailable");
+    }
+
+    log.info("AUTH", `\x1b[32mUsing ${providerId} account: ${credentials.connectionName}\x1b[0m`);
+
+    const refreshedCredentials = await checkAndRefreshToken(providerId, credentials);
+
+    const result = await handleFetchCore({
+      url: targetUrl,
+      format,
+      maxCharacters,
+      provider: resolvedProvider.id,
+      providerConfig,
+      credentials: refreshedCredentials,
+      log,
+      onCredentialsRefreshed: async (newCreds) => {
+        await updateProviderCredentials(credentials.connectionId, {
+          accessToken: newCreds.accessToken,
+          refreshToken: newCreds.refreshToken,
+          providerSpecificData: newCreds.providerSpecificData,
+          testStatus: "active"
+        });
+      },
+      onRequestSuccess: async () => {
+        await clearAccountError(credentials.connectionId, credentials);
+      }
+    });
+
+    if (result.success) {
+      return new Response(JSON.stringify(result.data), {
+        headers: { "Content-Type": "application/json", "Access-Control-Allow-Origin": "*" }
+      });
+    }
+
+    const { shouldFallback } = await markAccountUnavailable(credentials.connectionId, result.status, result.error, providerId);
+
+    if (shouldFallback) {
+      log.warn("AUTH", `Account ${credentials.connectionName} unavailable (${result.status}), trying fallback`);
+      excludeConnectionIds.add(credentials.connectionId);
+      lastError = result.error;
+      lastStatus = result.status;
+      continue;
+    }
+
+    return errorResponse(result.status || HTTP_STATUS.BAD_GATEWAY, result.error || "Fetch failed");
+  }
+}
--- a/src/sse/handlers/search.js
+++ b/src/sse/handlers/search.js
@@ -0,0 +1,204 @@
+import {
+  getProviderCredentials,
+  markAccountUnavailable,
+  clearAccountError,
+  extractApiKey,
+  isValidApiKey,
+} from "../services/auth.js";
+import { getSettings, getCombos } from "@/lib/localDb";
+import { AI_PROVIDERS, resolveProviderId } from "@/shared/constants/providers.js";
+import { handleSearchCore } from "open-sse/handlers/search/index.js";
+import { errorResponse, unavailableResponse } from "open-sse/utils/error.js";
+import { HTTP_STATUS } from "open-sse/config/runtimeConfig.js";
+import * as log from "../utils/logger.js";
+import { updateProviderCredentials, checkAndRefreshToken } from "../services/tokenRefresh.js";
+import { handleComboChat, getComboModelsFromData } from "open-sse/services/combo.js";
+
+/**
+ * Handle web search request for the SSE/Next.js server.
+ * Provider IS the model (no model field). Mirrors handleEmbeddings auth + fallback flow.
+ *
+ * @param {Request} request
+ */
+export async function handleSearch(request) {
+  let body;
+  try {
+    body = await request.json();
+  } catch {
+    log.warn("SEARCH", "Invalid JSON body");
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, "Invalid JSON body");
+  }
+
+  const url = new URL(request.url);
+  // Accept either `provider` or `model` (UI sends `model` since provider IS the model for webSearch)
+  const providerInput = body.provider || body.model;
+  const query = body.query;
+
+  log.request("POST", `${url.pathname} | ${providerInput}`);
+
+  // Log API key (masked)
+  const apiKey = extractApiKey(request);
+  if (apiKey) {
+    log.debug("AUTH", `API Key: ${log.maskKey(apiKey)}`);
+  } else {
+    log.debug("AUTH", "No API key provided (local mode)");
+  }
+
+  // Enforce API key if enabled in settings
+  const settings = await getSettings();
+  if (settings.requireApiKey) {
+    if (!apiKey) {
+      log.warn("AUTH", "Missing API key (requireApiKey=true)");
+      return errorResponse(HTTP_STATUS.UNAUTHORIZED, "Missing API key");
+    }
+    const valid = await isValidApiKey(apiKey);
+    if (!valid) {
+      log.warn("AUTH", "Invalid API key (requireApiKey=true)");
+      return errorResponse(HTTP_STATUS.UNAUTHORIZED, "Invalid API key");
+    }
+  }
+
+  if (!providerInput || typeof providerInput !== "string") {
+    log.warn("SEARCH", "Missing provider/model");
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, "Missing required field: provider (or model)");
+  }
+
+  if (!query || typeof query !== "string" || !query.trim()) {
+    log.warn("SEARCH", "Missing query");
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, "Missing required field: query");
+  }
+
+  // Combo expansion: providerInput may be a combo name → run fallback/round-robin across providers
+  const combos = await getCombos();
+  const comboModels = getComboModelsFromData(providerInput, combos);
+  if (comboModels) {
+    const comboStrategies = settings.comboStrategies || {};
+    const comboStrategy = comboStrategies[providerInput]?.fallbackStrategy || settings.comboStrategy || "fallback";
+    log.info("SEARCH", `Combo "${providerInput}" with ${comboModels.length} providers (strategy: ${comboStrategy})`);
+    return handleComboChat({
+      body,
+      models: comboModels,
+      handleSingleModel: (b, m) => handleSingleProviderSearch(b, m, request, apiKey, settings),
+      log,
+      comboName: providerInput,
+      comboStrategy
+    });
+  }
+
+  return handleSingleProviderSearch(body, providerInput, request, apiKey, settings);
+}
+
+async function handleSingleProviderSearch(body, providerInput, request, apiKey, settings) {
+  const query = body.query;
+  const providerId = resolveProviderId(providerInput);
+  const resolvedProvider = AI_PROVIDERS[providerId];
+
+  if (!resolvedProvider) {
+    log.warn("SEARCH", "Unknown provider", { provider: providerInput });
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, `Unknown provider: ${providerInput}`);
+  }
+
+  const providerConfig = resolvedProvider.searchConfig;
+  const supportsSearch = !!providerConfig || !!resolvedProvider.searchViaChat;
+
+  if (!supportsSearch) {
+    log.warn("SEARCH", "Provider does not support web search", { provider: providerId });
+    return errorResponse(HTTP_STATUS.BAD_REQUEST, `Provider ${providerId} does not support web search`);
+  }
+
+  if (providerInput !== providerId) {
+    log.info("ROUTING", `${providerInput} → ${providerId}`);
+  } else {
+    log.info("ROUTING", `Provider: ${providerId}`);
+  }
+
+  // Sanitized body forwarded to core
+  const coreBody = {
+    query: query.trim(),
+    provider: providerId,
+    max_results: body.max_results,
+    search_type: body.search_type,
+    country: body.country,
+    language: body.language,
+    time_range: body.time_range,
+    offset: body.offset,
+    domain_filter: body.domain_filter,
+    content_options: body.content_options,
+    provider_options: body.provider_options
+  };
+
+  // No-auth providers (e.g. searxng) bypass credential lookup
+  if (resolvedProvider.noAuth) {
+    log.info("AUTH", `\x1b[32m${providerId} no-auth mode\x1b[0m`);
+    const result = await handleSearchCore({
+      body: coreBody,
+      provider: resolvedProvider,
+      providerConfig,
+      credentials: null,
+      log
+    });
+    if (result.success) return result.response;
+    return result.response;
+  }
+
+  // Credential + fallback loop
+  const excludeConnectionIds = new Set();
+  let lastError = null;
+  let lastStatus = null;
+
+  while (true) {
+    const credentials = await getProviderCredentials(providerId, excludeConnectionIds);
+
+    if (!credentials || credentials.allRateLimited) {
+      if (credentials?.allRateLimited) {
+        const errorMsg = lastError || credentials.lastError || "Unavailable";
+        const status = lastStatus || Number(credentials.lastErrorCode) || HTTP_STATUS.SERVICE_UNAVAILABLE;
+        log.warn("SEARCH", `[${providerId}] ${errorMsg} (${credentials.retryAfterHuman})`);
+        return unavailableResponse(status, `[${providerId}] ${errorMsg}`, credentials.retryAfter, credentials.retryAfterHuman);
+      }
+      if (excludeConnectionIds.size === 0) {
+        log.error("AUTH", `No credentials for provider: ${providerId}`);
+        return errorResponse(HTTP_STATUS.BAD_REQUEST, `No credentials for provider: ${providerId}`);
+      }
+      log.warn("SEARCH", "No more accounts available", { provider: providerId });
+      return errorResponse(lastStatus || HTTP_STATUS.SERVICE_UNAVAILABLE, lastError || "All accounts unavailable");
+    }
+
+    log.info("AUTH", `\x1b[32mUsing ${providerId} account: ${credentials.connectionName}\x1b[0m`);
+
+    const refreshedCredentials = await checkAndRefreshToken(providerId, credentials);
+
+    const result = await handleSearchCore({
+      body: coreBody,
+      provider: resolvedProvider,
+      providerConfig,
+      credentials: refreshedCredentials,
+      log,
+      onCredentialsRefreshed: async (newCreds) => {
+        await updateProviderCredentials(credentials.connectionId, {
+          accessToken: newCreds.accessToken,
+          refreshToken: newCreds.refreshToken,
+          providerSpecificData: newCreds.providerSpecificData,
+          testStatus: "active"
+        });
+      },
+      onRequestSuccess: async () => {
+        await clearAccountError(credentials.connectionId, credentials);
+      }
+    });
+
+    if (result.success) return result.response;
+
+    const { shouldFallback } = await markAccountUnavailable(credentials.connectionId, result.status, result.error, providerId);
+
+    if (shouldFallback) {
+      log.warn("AUTH", `Account ${credentials.connectionName} unavailable (${result.status}), trying fallback`);
+      excludeConnectionIds.add(credentials.connectionId);
+      lastError = result.error;
+      lastStatus = result.status;
+      continue;
+    }
+
+    return result.response;
+  }
+}