9Router - FREE AI Router & Token Saver

9Router - FREE AI Router & Token Saver

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.

🚀 Quick Start • 💡 Features • 📖 Setup • 🌐 Website

🇧🇷 Português (Brasil) • 🇻🇳 Tiếng Việt • 🇨🇳 中文 • 🇯🇵 日本語 • 🇷🇺 Русский • 🇹🇭 ไทย • 🇮🇷 فارسی • 🇮🇩 Indonesia • 🇪🇸 Español • 🇫🇷 Français

🔀 This is a feature-enhanced fork of decolua/9router, adding a managed V2Ray/Xray proxy (v2go), DeepSeek Web (DS2API) sidecar, rotating proxy pools, Genspark/Gemini web-cookie providers, external tunnel URL, and more. Distributed via GitHub Releases (not npm). See ⭐ Fork Features below.

🤔 Why 9Router?

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, Cursor, Cline, any CLI tool

🔄 How It Works

┌─────────────┐
│  Your CLI   │  (Claude Code, Codex, OpenClaw, Cursor, Cline...)
│   Tool      │
└──────┬──────┘
       │ http://localhost:20128/v1
       ↓
┌─────────────────────────────────────────────┐
│           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, GitHub Copilot
       │   ↓ quota exhausted
       ├─→ [Tier 2: CHEAP] GLM ($0.6/1M), MiniMax ($0.2/1M)
       │   ↓ budget limit
       └─→ [Tier 3: FREE] Kiro, OpenCode Free, Vertex ($300 credits)

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

⭐ Fork Features

Additions in this fork (vibecoder11200/9router) on top of upstream. All are optional and ship disabled by default.

Feature	What it adds	Where to enable
🛰️ V2Ray Proxy (v2go)	Managed local Xray-core client that turns V2Ray share links (VLESS/VMess/Trojan/SS) from v2go and any other subscription into a SOCKS5/HTTP proxy 9Router can route through. Multi-subscription sync with per-sub interval/retention/traffic display, in-dashboard binary updates/downgrades (stable-latest check, auto-restart with rollback), per-server latency testing, zero-downtime blue-green auto-rotation (flaky-node + edge-banned-IP quarantine), and a Model Proxy Filter that finds the configs a given model actually works through. Bundles the Xray-core binary (auto-download per OS/arch). Creates a managed Proxy Pool you can assign to any connection. (v0.6.0+)	Dashboard → V2Ray Proxy
🐬 DeepSeek Web (DS2API)	Runs a local Go sidecar that turns your DeepSeek Web session into an OpenAI-compatible endpoint. Managed start/stop/install/update, per-account proxies + rotating proxy groups (round-robin/random/failover). Engine pulled from `vibecoder11200/ds2api` `v4.6.2-rotation`.	Dashboard → DeepSeek Web
🔀 Proxy Pools & Rotating Groups	Single-proxy pools or rotating groups (many proxies + optional "direct" server-IP slot). Per-request rotation: on-error (LRU) / round-robin / random. All protocols (http, https, socks5/5h/4/4a). Batch import. `strictProxy` fail-hard. Auto-cooldown (60s rate-limit, 30s 5xx). Bind to any provider connection.	Dashboard → Proxy Pools
🌐 No-auth provider rotation	Free no-auth providers (OpenCode Free, mimo-free…) can be bound to a rotating pool group from their provider page — set Rotation Strategy to round-robin/random (needs ≥2 active pools) to spread requests across IPs.	Provider page → Proxy / Rotation card
🤖 Genspark Web	Cookie-based Genspark Copilot MOA backend. Chat + image generation (`COPILOT_MOA_IMAGE`). Append `-search` to any model for web grounding. Prefix `genspark-web/` (`gspark`).	Dashboard → Providers → Genspark Web
♊ Gemini Web	Cookie-based `gemini.google.com` (internal `StreamGenerate` RPC). Cookie pool up to 5, round-robin, 15-min health checks, auto-disable dead cookies. LLM + image + video + audio. Prefix `gemini-web/` (`gweb`).	Dashboard → Providers → Gemini Web
🐟 TOTU AI Auto-Fetch (Lấy acc)	One-click account farming for the TOTU AI free NewAPI gateway: creates a temp mail.tm mailbox, captures the email OTP, registers + logs in, and saves the `sk-` key (with the dashboard login token) as a provider connection — plus a per-account $ balance view (shared with TokenRouter). Optional scheduler auto-fetches on an interval (default off; 15/30/60 min). (v0.6.29+)	Dashboard → Providers → TOTU AI → Lấy acc
🔗 External Tunnel URL	Register a tunnel the app does not manage (e.g. `cloudflared` via systemd, or any reverse proxy). Combined with Allow dashboard access via tunnel, local-only actions (DS2API install/start/stop, tunnel controls, Headroom, MITM) run over that tunnel after login. Setting `externalTunnelUrl`.	Dashboard → Endpoint → External tunnel URL

Plus everything from upstream (regularly merged): PXPipe multimodal token saver, Grok CLI, Perplexity Agent API, Featherless, self-hosted STT/TTS/embedding providers, Headroom extras — all documented in their sections below.

📖 How the two proxy-group systems differ

This fork has two independent proxy-group systems. They are easy to confuse:

9Router Proxy Pools (Dashboard → Proxy Pools) — 9Router's own. Modes: on-error / round-robin / random. Applies to any provider connection. Cools down failing entries and tries another entry on the same account before account fallback. Code: src/lib/network/proxyRotation.js.
DS2API proxy groups (Dashboard → DeepSeek Web) — managed inside the DS2API Go sidecar and surfaced through the dashboard. Modes: round-robin / random / failover (+ sticky count). Applies only to DeepSeek Web accounts. Code: temp/ds2api/internal/config.

⚡ Quick Start

1. Install globally:

This fork is distributed via GitHub Releases, not the npm registry. Pick the one-liner for your platform (Node.js >= 18 required):

# macOS / Linux / WSL
curl -fsSL https://github.com/vibecoder11200/9router/raw/master/install.sh | bash

# Windows (PowerShell)
powershell -c "irm https://github.com/vibecoder11200/9router/raw/master/install.ps1 | iex"

# …or install the latest release tarball directly:
npm install -g https://github.com/vibecoder11200/9router/releases/latest/download/9router.tgz

9router

🎉 Dashboard opens at http://localhost:20128

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

Dashboard → Providers → Connect Kiro AI (~50 credits/month free: Claude 4.5 + GLM-5 + MiniMax) or OpenCode Free (no auth) → Done!

3. Use in your CLI tool:

Claude Code/Codex/OpenClaw/Cursor/Cline Settings:
  Endpoint: http://localhost:20128/v1
  API Key: [copy from dashboard]
  Model: kr/claude-sonnet-4.5

That's it! Start coding with FREE AI models.

Alternative: run from source (this repository):

This repository package is private (9router-app), so source/Docker execution is the expected local development path.

cp .env.example .env
npm install
PORT=20128 NEXT_PUBLIC_BASE_URL=http://localhost:20128 npm run dev

Production mode:

npm run build
PORT=20128 HOSTNAME=0.0.0.0 NEXT_PUBLIC_BASE_URL=http://localhost:20128 npm run start

Default URLs:

Dashboard: http://localhost:20128/dashboard
OpenAI-compatible API: http://localhost:20128/v1

🔄 Update / Nâng cấp

This fork ships via GitHub Releases (not npm). The dashboard checks for new releases automatically on the fork's GitHub Releases and shows an "↑ New version available" badge in the sidebar when one is found — click it to copy the upgrade command and shut the server down safely.

To upgrade manually, re-run the install one-liner (it always pulls latest):

# macOS / Linux / WSL
curl -fsSL https://github.com/vibecoder11200/9router/raw/master/install.sh | bash

# Windows (PowerShell)
powershell -c "irm https://github.com/vibecoder11200/9router/raw/master/install.ps1 | iex"

# …or install the latest release tarball directly:
npm install -g https://github.com/vibecoder11200/9router/releases/latest/download/9router.tgz

The CLI launcher also checks for updates on start and prints the exact upgrade command when a newer release exists (run 9router and watch the menu).

Note: Do not run npm i -g 9router / npm i -g 9router@latest — that installs the upstream decolua package from npm, not this fork. Always use the GitHub Releases tarball URL or the install scripts above.

Video Guides


_{🇻🇳 Tiết kiệm chi phí LLM cho OpenClaw với 9Router by Mì AI}	_{🇻🇳 Cài Đặt OpenClaw Free Từ A-Z + 9Router by Mai Gia}

_{🇻🇳 Setup OpenClaw + 9Router: Bot Zalo AI Tự Động A-Z by tuanminhhole}	_{🇺🇸 9Router + Claude Code FREE Setup by Build AI With Hamid}

_{🇺🇸 Claude Code FREE Forever — Unlimited Models by Build AI With Hamid}	_{🇺🇸 Claude CLI Free Setup with 9Router by CodeVerse Soban}

_{🇵🇰 9Router + Claude Code FREE Unlimited Setup by Build AI With Hamid}	_{🇺🇸 FREE OpenClaw + Claude Opus 4.6 by Build AI With Hamid}

_{🇺🇸 9Router + Claude Code FREE Setup by Build AI With Hamid}	_{🇮🇩 Koding 24 Jam Anti Rate Limit! Hemat Token AI 65% by Krisswuh}

_{🇮🇩 Deploy 9Router di Hugging Face GRATIS Non-Stop by Krisswuh}	_{🇮🇷 این شکلی از هر API ای استفاده کن برای هوش مصنوعی by Matin SenPai}

🎬 Made a video about 9Router? Submit a Pull Request adding your video to this section — we'll merge it!

🛠️ Supported CLI Tools

9Router works seamlessly with all major AI coding tools:

Claude-Code	OpenClaw	Codex	OpenCode	Cursor	Antigravity
Cline	Continue	Droid	Roo	Copilot	Kilo Code
OpenDesign	jcode	Grok Build	Devin CLI	DeepSeek TUI	Qwen Code

🌐 Supported Providers

🔐 OAuth Providers

Claude-Code

Antigravity

Codex

GitHub

Cursor

Kimchi

🆓 Free Providers

Kiro AI
_{Claude 4.5 + GLM-5 + MiniMax
50 credits/month free}

OpenCode Free
_{No auth • Auto-fetch models
Free (model list varies)}

Vertex AI
_{Gemini 3 Pro + GLM-5 + DeepSeek
$300 credits free}

Note: iFlow, Qwen Code and Gemini CLI free tiers were discontinued in 2026. Use Kiro / OpenCode Free / Vertex instead.

Kiro AI moved to a paid model in Sep 2025 — the free tier is now capped at 50 credits/month (plus 500 trial credits for new accounts in the first 30 days). Paid tiers: Pro $20/mo (1,000 credits), Pro+ $40/mo (2,000), Pro Max $100/mo (5,000), Power $200/mo (10,000). OpenCode Free model list fluctuates over time (some models free only for limited promos) — subject to change without notice. Vertex AI: the $300 free credit for new GCP accounts is still valid, but since Mar 2026 the Gemini API endpoint no longer consumes these credits — call the Vertex AI Studio endpoint instead.

🍪 Web-Cookie Providers · fork

Authenticate with a browser session cookie instead of an API key — turns web-only AI into an OpenAI-compatible endpoint. Added by this fork.

Provider	Prefix	What you get
Gemini Web	`gemini-web/` (`gweb`)	`gemini.google.com` via internal RPC. LLM + image + video + audio. Cookie pool (up to 5, round-robin, 15-min health checks, auto-disable dead cookies).
Genspark Web	`genspark-web/` (`gspark`)	Genspark Copilot MOA chat + image generation (`COPILOT_MOA_IMAGE`). Append `-search` to any model for web grounding.
DeepSeek Web	`ds2api/`	Your DeepSeek Web session, via a managed local sidecar. See ⭐ Fork Features.

Setup: open the provider in Dashboard → Providers, paste the session cookie (JSON from a cookie editor, or the bare session_id value), and the models appear automatically.

🔑 API Key Providers (40+)

_OpenRouter	_GLM	_Kimi	_MiniMax	_OpenAI	_Anthropic
_Gemini	_DeepSeek	_Groq	_xAI	_Mistral	_Perplexity
_{Together AI}	_Fireworks	_Cerebras	_Cohere	_NVIDIA	_SiliconFlow

...and 20+ more providers including Grok CLI (OAuth), Perplexity Agent API, Featherless, Cloudflare AI, Nebius, Chutes, Hyperbolic, Venice AI, TokenRouter, OrcaRouter, TOTU AI, and custom OpenAI/Anthropic compatible endpoints

🏠 Self-hosted Providers

For speech and embeddings served from your own machine — whisper.cpp, faster-whisper, Speaches, Kokoro-FastAPI, openedai-speech, llama.cpp/llama-server, vLLM, Infinity, text-embeddings-inference, or anything else that speaks the OpenAI shape.

Provider	Endpoint used	Typical server
Self-hosted STT	`/v1/audio/transcriptions`	whisper.cpp, faster-whisper
Self-hosted TTS	`/v1/audio/speech`	Kokoro-FastAPI, openedai-speech
Self-hosted Embedding	`/v1/embeddings`	llama-server, vLLM, Infinity

Every other speech provider is a named cloud service with a fixed endpoint. These three read their address from each connection, so one provider can front several machines and load-balance across them like any other.

Set it on the connection as providerSpecificData.baseUrl:

Provider	Give it	Result
Self-hosted STT	the full URL — `http://host:8080/v1/audio/transcriptions`	used as-is
Self-hosted TTS	the server root — `http://host:8880`	`+ /v1/audio/speech`
Self-hosted Embedding	the OpenAI base, `/v1` included — `http://host:8080/v1`	`+ /embeddings`

Mind the /v1 on embeddings. The adapter appends /embeddings, so http://host:8080 resolves to http://host:8080/embeddings and misses the OpenAI route — llama-server answers 501. Give it the same base URL an OpenAI client would use. A full .../v1/embeddings is also accepted, so a value pasted from a curl example works too.

The API key is not checked by most local servers, but the field must be non-empty: it is what gives the connection a credentials record, and baseUrl lives there. Any placeholder works.

Self-hosted Embedding has no cloud fallback by design — a connection saved without a baseUrl is reported as a configuration error rather than quietly falling back to api.openai.com, which would send your input text and API key to a third party under a provider named "Self-hosted".

💡 Key Features

Feature	What It Does	Why It Matters
🚀 RTK Token Saver (RTK ⭐40K)	Compress tool outputs (`git diff`, `grep`, `ls`, `tree`...) before sending to LLM	Save 20-40% input tokens per request
🧠 Headroom Token Saver (Headroom)	Optional external `/v1/compress` proxy before provider routing	Save more context tokens without changing clients
🖼️ PXPipe Token Saver	In-process multimodal compression — re-renders Claude-format context as dense images (Anthropic bills images by pixels, not text length)	Save context tokens on long Claude requests
🪨 Caveman Mode (Caveman ⭐52K)	Inject caveman-speak prompt → LLM replies terse, technical substance preserved	Save up to 65% output tokens
🐴 Ponytail (Ponytail)	Inject "lazy senior dev" prompt → LLM writes minimal, YAGNI-first code (Lite/Full/Ultra)	Fewer output tokens, less refactoring
🛰️ V2Ray Proxy (v2go) · fork	Managed Xray-core client → SOCKS5/HTTP proxy from V2Ray subscriptions (multi-sub sync, in-dashboard binary updates)	Premium-grade proxies for any provider, free
🐬 DeepSeek Web (DS2API) · fork	Local Go sidecar turns your DeepSeek Web session into an OpenAI endpoint	Use DeepSeek Web from any CLI tool
🔀 Proxy Pools & Rotating Groups · fork	Single-proxy pools or rotating groups (on-error/round-robin/random + direct slot)	Spread load, beat IP rate-limits
🤖 Web-Cookie Providers · fork	Genspark (MOA + image), Gemini Web (multimodal, cookie pool)	Access web-only AI in any CLI tool
🎯 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 ↔ 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
📝 Request Logging	Debug mode with full request/response logs	Troubleshoot issues easily
💾 Cloud Sync	Sync config across devices	Same setup everywhere
📊 Usage Analytics	Track tokens, cost, trends over time	Optimize spending
🌐 Deploy Anywhere	Localhost, VPS, Docker, Cloudflare Workers	Flexible deployment options

Set X-9Router-Token-Saver: off to bypass all token savers for one chat request.

📖 Feature Details

🚀 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)

🧠 Headroom Token Saver

Headroom is optional and runs separately. 9Router calls Headroom's local /v1/compress endpoint, then keeps normal routing, fallback, auth, and usage tracking:

Client → 9Router → Headroom /v1/compress → 9Router → provider

Local setup:

pip install "headroom-ai[proxy]"
headroom proxy --port 8787

Enable in Dashboard → Endpoint → Token Saver → Headroom. Default URL: http://localhost:8787 (override with HEADROOM_URL).

Optional extras (install from the same Headroom card in the dashboard):

code — tree-sitter AST-based code compression.
ml — Kompress-v2 HuggingFace model compression.

The dashboard auto-detects installed extras via pip list, and offers one-click install/uninstall with a live log. Other extras (image, voice, otel, …) aren't tracked since they don't help token compression.

Docker examples:

# Headroom service in same Docker network
http://headroom:8787

# Headroom running on host machine
http://host.docker.internal:8787

If Headroom is down or returns an error, 9Router fails open and sends the original request.

🖼️ PXPipe Token Saver

PXPipe is a multimodal compressor: it re-renders dense Claude-format text context as compact images. Anthropic bills images by pixels (pixels/750) rather than encoded text length, so a long context can cost fewer tokens as an image than as text.

In-process — runs as a library inside 9Router (no separate daemon/port). The npm package is installed on first enable.
Claude-only — only transforms Claude-format requests above a size threshold (pxpipeMinChars, default 25000 chars).
Fail-open — any error/timeout leaves the request untouched.
Default off — enable in Dashboard → Token Saver (the pxpipeEnabled toggle). Stats and a health check live under Dashboard → Pxpipe.

Stacks with RTK (which runs first and strips agentic noise) and Headroom (external text compression).

🐴 Ponytail (Lazy Senior Dev)

Ponytail injects a "lazy senior dev" system prompt into every request, biasing the LLM toward minimal, YAGNI-first code — deletion over addition, stdlib over new deps, one-liners over abstractions. Adapted from DietrichGebert/ponytail.

Lite — Build what's asked, name the lazier alternative.
Full — YAGNI ladder enforced: stdlib → native → existing deps → one-liner → minimal code.
Ultra — YAGNI extremist: deletion first, ship the one-liner, challenge the rest of the requirement in the same response.

Without Ponytail: verbose code, extra abstractions, "just in case" scaffolding
With Ponytail:    shortest working diff, no unrequested abstractions, fewer tokens

Never trades away: input validation, error handling that prevents data loss, security, accessibility, or anything explicitly requested. Enable in Dashboard → Endpoint → Ponytail. Stacks with Caveman (output terseness) and RTK (input compression).

🎯 Smart 3-Tier Fallback

Create combos with automatic fallback:

Combo: "my-coding-stack"
  1. cc/claude-opus-5          (your subscription)
  2. glm/glm-4.7               (cheap backup, $0.6/1M)
  3. kr/claude-sonnet-4.5      (free fallback)

→ Auto switches when quota runs out or errors occur

📊 Real-Time Quota Tracking

Token consumption per provider
Reset countdown (5-hour, daily, weekly)
Cost estimation for paid tiers
Monthly spending reports

🔄 Format Translation

Seamless translation between formats:

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

👥 Multi-Account Support

Add multiple accounts per provider
Auto round-robin or priority-based routing
Fallback to next account when one hits quota

🔄 Auto Token Refresh

OAuth tokens automatically refresh before expiration
No manual re-authentication needed
Seamless experience across all providers

🎨 Custom Combos

Create unlimited model combinations
Mix subscription, cheap, and free tiers
Name your combos for easy access
Share combos across devices with Cloud Sync

📝 Request Logging

Enable debug mode for full request/response logs
Track API calls, headers, and payloads
Troubleshoot integration issues
Export logs for analysis

💾 Cloud Sync

Sync providers, combos, and settings across devices
Automatic background sync
Secure encrypted storage
Access your setup from anywhere

Cloud Runtime Notes

Prefer server-side cloud variables in production:
- BASE_URL (internal callback URL used by sync scheduler)
- CLOUD_URL (cloud sync endpoint base)
NEXT_PUBLIC_BASE_URL and NEXT_PUBLIC_CLOUD_URL are still supported for compatibility/UI, but server runtime now prioritizes BASE_URL/CLOUD_URL.
Cloud sync requests now use timeout + fail-fast behavior to avoid UI hanging when cloud DNS/network is unavailable.

📊 Usage Analytics

Track token usage per provider and model
Cost estimation and spending trends
Monthly reports and insights
Optimize your AI spending

💡 IMPORTANT - Understanding Dashboard Costs:

The "cost" displayed in Usage Analytics is for tracking and comparison purposes only. 9Router itself never charges you anything. You only pay providers directly (if using paid services).

Example: If your dashboard shows "$290 total cost" while using Kiro models, this represents what you would have paid using paid APIs directly. Your actual cost = $0 (Kiro free tier: ~50 credits/mo).

Think of it as a "savings tracker" showing how much you're saving by using free models or routing through 9Router!

🌐 Deploy Anywhere

💻 Localhost - Default, works offline
☁️ VPS/Cloud - Share across devices
🐳 Docker - One-command deployment
🚀 Cloudflare Workers - Global edge network

💰 Cost & Strategy

Pricing at a glance

Tier	Provider	Cost	Quota Reset	Best For
🚀 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
	GitHub Copilot	$10-19/mo	Monthly	GitHub users
	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	50 credits/mo	Claude 4.5 + GLM-5 + MiniMax free (paid tiers above)
	OpenCode Free	$0	Varies*	No auth, auto-fetch models (list changes over time)
	Vertex AI	$300 credits	New GCP accounts	Gemini 3 Pro + DeepSeek + GLM-5 (use Vertex AI Studio endpoint for free credits)

Billing reality

✅ 9Router = FREE forever (open source, never charges, no invoices, no credit card).
✅ You pay providers directly — subscriptions on their websites, or API fees. 9Router just routes.
✅ Dashboard "cost" is a savings tracker, not a bill. It shows what you would have paid using paid APIs — e.g. "$290 displayed" while using Kiro free tier means you saved $290, actual payment $0.
❌ FREE providers have free-tier limits (Kiro ~50 credits/mo, OpenCode/Vertex per their terms). iFlow/Qwen/Gemini CLI free tiers were discontinued in 2026.

Example combos

Combo	Layers (fallback top → bottom)	Monthly cost
free-forever	`kr/claude-sonnet-4.5` → `kr/glm-5` → `oc/<auto>` (OpenCode Free)	$0
maximize-claude	`cc/claude-opus-5` (subscription) → `glm/glm-5.1` (cheap) → `kr/claude-sonnet-4.5` (free fallback)	~$25
always-on	`cc/claude-opus-5` → `cx/gpt-5.5` → `glm/glm-5.1` → `minimax/MiniMax-M2.7` → `kr/claude-sonnet-4.5`	$30-220
openclaw-free	`kr/claude-sonnet-4.5` → `kr/glm-5` → `kr/MiniMax-M2.5` — free AI in WhatsApp/Telegram/Slack/Discord/iMessage/Signal	$0

💡 Pro Tip: RTK + Kiro + OpenCode Free = $0 cost + 20-40% token savings.

❓ Frequently Asked Questions

📊 Why does my dashboard show high costs? Will I be charged?

The dashboard "cost" is a savings tracker, not a bill — it shows what you would have paid using paid APIs. 9Router never charges you (open source, no invoices, no credit card). You only pay providers directly (subscriptions/API fees). Example: "$290 displayed" while using Kiro free tier = you saved $290, actual payment $0. See 💰 Cost & Strategy for full pricing, free-tier limits, and example combos.

🆓 Are FREE providers really free? Which ones still work?

Yes, within free-tier limits — Kiro AI (~50 credits/mo + 500 trial credits for new accounts in first 30 days), OpenCode Free (no-auth, model list fluctuates), Vertex AI ($300 credits for new GCP accounts — use the Vertex AI Studio endpoint since the Gemini API stopped consuming credits in Mar 2026). 9Router just routes; no catch, no future billing.

Discontinued (don't use): ❌ iFlow (paid since 2026) · ❌ Qwen Code (Alibaba discontinued free OAuth 2026-04-15) · ❌ Gemini CLI (Google shut down 2026-06-18, replaced by Antigravity CLI).

📖 Setup Guide

🔌 Connect a provider

All providers connect from Dashboard → Providers. Point your CLI tool at http://localhost:20128/v1 (API key from the dashboard), then use the model prefix.

Tier	Provider	How to connect	Model examples
💳 Subscription	Claude Code	OAuth login → auto token refresh	`cc/claude-opus-5`, `cc/claude-sonnet-5`
💳	Codex	OAuth (port 1455)	`cx/gpt-5.6-sol`, `cx/gpt-5.5`
💳	GitHub Copilot	OAuth via GitHub (monthly reset)	`gh/gpt-5.4`, `gh/claude-opus-4.7`, `gh/gemini-3.1-pro-preview`
💳	Cursor IDE	OAuth login	`cu/claude-4.6-opus-max`, `cu/gpt-5.3-codex`
💰 Cheap	GLM	API key from Zhipu AI — Coding Plan = 3× quota at 1/7 cost, resets daily 10 AM	`glm/glm-5.1`, `glm/glm-4.7`
💰	MiniMax	API key from MiniMax — cheapest for long context	`minimax/MiniMax-M2.7`
💰	Kimi	API key from Moonshot AI — $9/mo flat for 10M tokens	`kimi/kimi-k2.5`, `kimi/kimi-k2.5-thinking`

Pro Tip: In any combo, order models Subscription → Cheap → Free so 9Router auto-falls through tiers when quota runs out. See 💰 Cost & Strategy for ready-made combos.

🆓 FREE providers (recommended)

Kiro AI · kr/ ~50 credits/mo free (500 trial for new accounts in first 30 days). Best free Claude.

Dashboard → Connect Kiro
→ AWS Builder ID / Google / GitHub

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/ No login — passthrough proxy. Fastest setup. Model list auto-fetched from opencode.ai/zen/v1/models (fluctuates over time).

Dashboard → Connect OpenCode Free
→ No login required
→ Use oc/<auto> in combos

Vertex AI · vertex/ $300 free credits for new GCP accounts (90 days). Use the Vertex AI Studio endpoint — the Gemini API endpoint stopped consuming credits in Mar 2026.

Dashboard → Connect Vertex AI
→ Upload GCP Service Account JSON

vertex/gemini-3.1-pro-preview
vertex/gemini-3-flash-preview
vertex-partner/glm-5-maas
vertex-partner/deepseek-v3.2-maas

🛰️ V2Ray Proxy (v2go) · fork

9Router manages a local Xray-core client that turns free V2Ray share links (VLESS/VMess/Trojan/Shadowsocks) into a SOCKS5/HTTP proxy 9Router can route any provider through. The config catalog auto-syncs from one or more subscriptions — v2go by default (~1,000+ working servers via a GitHub Actions pipeline) — and any other V2Ray subscription URL you add.

Setup

Dashboard → V2Ray Proxy
  → Install Xray-core   (downloads the official binary; updates + downgrades
                         with auto-restart and rollback stay in-dashboard)
  → Add subscriptions   (v2go is pre-configured as the "Default" subscription;
                         each sub gets its own schedule + retention)
  → Sync configs        (per-sub "Sync Now" or "Sync All" — VLESS/VMess/Trojan/SS)
  → Pick a server       (filter by protocol; run a latency test)
  → Start               (launches the local SOCKS5 + HTTP proxy)
  → Assign the pool     (a managed Proxy Pool "V2Ray Proxy (v2go)" is created
                         automatically — bind it to any provider connection)

Once started, a managed Proxy Pool named "V2Ray Proxy (v2go)" appears in Dashboard → Proxy Pools. Assign it to any provider connection (or to a no-auth provider's Proxy / Rotation card) and that connection's traffic routes through the active SOCKS proxy.

Features

Multi-subscription sync — each subscription has its own enable toggle, sync interval, and keep-dropped-servers retention; per-row traffic + expiry display when the provider reports a subscription-userinfo header; a failing subscription never affects the others.
Server selection — protocol filters, per-server latency testing, source-subscription badges, and a Deleted servers expander (restore or permanently delete).
Binary updates — in-dashboard Xray-core updates/downgrades: stable-latest check (pre-releases labeled in the picker, never trigger the update badge), auto-restart after update with automatic rollback if the new binary can't start.
Auto-rotation (optional) — when the active server dies, 9Router promotes the next-best server automatically.
Health checks — periodic liveness checks (default every 10 min).
Share-link parser — a faithful JS port of v2go's converter.go: handles VLESS, VMess, Trojan, Shadowsocks, Hysteria2 with REALITY/TLS/WebSocket/gRPC/XHTTP transports, including the XHTTP host-safety guard that prevents Xray crashes.

Engine / settings

The Xray-core binary is pinned to v26.3.27 (MPL-2.0) and auto-downloaded on first use — override the tag with the XRAY_VERSION env var. The rest is configured from the dashboard (stored as settings, not env vars):

Setting	Default	Purpose
`xrayEnabled`	`false`	Master on/off for the V2Ray proxy
`xrayAutoStart`	`false`	Start the proxy when 9Router boots
`xrayAutoRotate`	`false`	Promote the next server when the active one dies
`xraySocksPort`	`10808`	Local SOCKS5 port
`xrayHttpPort`	`10809`	Local HTTP proxy port
`xraySubscriptionUrl`	v2go `AllConfigsSub.txt`	Legacy — migrated to the "Default" subscription on upgrade; still the default URL offered to new subscriptions
`xraySyncIntervalMin`	`60`	Default sync interval applied to new subscriptions (per-sub intervals live on each subscription)
`xrayHealthCheckIntervalMin`	`10`	Liveness check interval (minutes)

🔀 Proxy Pools & Rotating Groups · fork

A proxy pool is either a single proxy or a rotating group of many proxies (plus an optional "direct" server-IP slot). Bind it to any provider connection so that connection's outbound traffic goes through the pool.

Create a pool

Dashboard → Proxy Pools → Create

  Type:
    • Single proxy  → one proxyUrl (http/https/socks5/socks5h/socks4/socks4a)
    • Rotating group → multiple entries + rotation mode

  Rotating group options:
    Rotation mode:
      • on-error  (default) — least-recently-used, skips the entry that just failed
      • round-robin — advance to the next entry every request
      • random    — uniform random per request
    Entries:   +proxy  (paste a proxy URL)
               +direct (server's own IP, no proxy)
    strictProxy: ☐  fail hard if the proxy errors (don't fall back to direct)

Batch import: paste a proxy list (protocol://user:pass@host:port or host:port:user:pass) to add many entries at once (deduped automatically).

Bind to a connection

Open a provider connection → Proxy → select the pool. Free no-auth providers (OpenCode Free, mimo-free) instead show a Proxy / Rotation card on the provider page: set Rotation Strategy to round-robin or random (needs ≥2 active pools) to spread requests across IPs.

How rotation behaves at runtime

On a rotatable error (408/429/rate-limit/quota/capacity/overloaded/5xx), the current entry is cooled down (60s for rate-limits, 30s for 5xx) and the next entry is tried on the same account.
Only when the whole group is exhausted does 9Router fall back to the next account/combo tier.
strictProxy = on disables that graceful fallback for the pool — a failing proxy fails the request instead of leaking your real IP.

🐬 DeepSeek Web (DS2API) · fork

9Router manages a local Go sidecar that turns your DeepSeek Web session into an OpenAI-compatible endpoint, so any CLI tool can use DeepSeek Web.

Setup

Dashboard → DeepSeek Web
  → Install engine   (downloads vibecoder11200/ds2api v4.6.2-rotation, ~one-time)
  → Add account      (paste your DeepSeek Web credentials)
  → Start engine
  → Enable           (toggles the ds2api provider connection + auto-aliases models)

Models are auto-aliased with the ds2api/ prefix on managed start (e.g. ds2api/deepseek-chat), so OpenAI clients work without the prefix too.

Per-account proxies & rotating groups

The DS2API sidecar has its own proxy-group system (separate from 9Router's Proxy Pools):

Dashboard → DeepSeek Web → Proxy groups (rotating)
  Strategy: round-robin | random | failover
  Sticky:   N   (requests before rotating — round-robin only, 1–1000)

Each account row → proxy mode: direct | fixed | group

round-robin — advance every N requests (sticky).
random — uniform per request.
failover — retry on the next proxy on transport error / 5xx / 408 / 429, replaying the request body.

Engine / env

The engine is pulled from the vibecoder11200/ds2api fork (release v4.6.2-rotation) which adds HTTP/HTTPS proxy support on top of upstream's socks5-only build. Override with:

Env var	Purpose
`DS2API_VERSION`	Engine release tag (default `v4.6.2-rotation`)
`DS2API_URL`	Override the sidecar loopback URL
`DS2API_ADMIN_KEY`	Override the auto-generated admin secret
`DS2API_CONFIG_PATH`	Sidecar config file location (default `${DATA_DIR}/ds2api/config.json`)

🔧 CLI Integration

Cursor IDE

Settings → Models → Advanced:
  OpenAI API Base URL: http://localhost:20128/v1
  OpenAI API Key: [from 9router dashboard]
  Model: cc/claude-opus-5

Or use combo: premium-coding

Claude Code

Edit ~/.claude/config.json:

{
  "anthropic_api_base": "http://localhost:20128/v1",
  "anthropic_api_key": "your-9router-api-key"
}

Codex CLI

export OPENAI_BASE_URL="http://localhost:20128"
export OPENAI_API_KEY="your-9router-api-key"

codex "your prompt"

OpenClaw

Option 1 — Dashboard (recommended):

Dashboard → CLI Tools → OpenClaw → Select Model → Apply

Option 2 — Manual: Edit ~/.openclaw/openclaw.json:

{
  "agents": {
    "defaults": {
      "model": {
        "primary": "9router/kr/claude-sonnet-4.5"
      }
    }
  },
  "models": {
    "providers": {
      "9router": {
        "baseUrl": "http://127.0.0.1:20128/v1",
        "apiKey": "sk_9router",
        "api": "openai-completions",
        "models": [
          {
            "id": "kr/claude-sonnet-4.5",
            "name": "Claude Sonnet 4.5 (Kiro Free)"
          }
        ]
      }
    }
  }
}

Note: OpenClaw only works with local 9Router. Use 127.0.0.1 instead of localhost to avoid IPv6 resolution issues.

Cline / Continue / RooCode

Provider: OpenAI Compatible
Base URL: http://localhost:20128/v1
API Key: [from dashboard]
Model: cc/claude-opus-5

🚀 Deployment

VPS Deployment

# Clone and install
git clone https://github.com/vibecoder11200/9router.git
cd 9router
npm install
npm run build

# Configure
export JWT_SECRET="your-secure-secret-change-this"
export INITIAL_PASSWORD="your-password"
export DATA_DIR="/var/lib/9router"
export PORT="20128"
export HOSTNAME="0.0.0.0"
export NODE_ENV="production"
export NEXT_PUBLIC_BASE_URL="http://localhost:20128"
export NEXT_PUBLIC_CLOUD_URL="https://9router.com"
export API_KEY_SECRET="endpoint-proxy-api-key-secret"
export MACHINE_ID_SALT="endpoint-proxy-salt"

# Start
npm run start

# Or use PM2
npm install -g pm2
pm2 start npm --name 9router -- start
pm2 save
pm2 startup

Docker

Published images (multi-platform linux/amd64 + linux/arm64):

Docker Hub: vibecoder11200/9router
GHCR: ghcr.io/vibecoder11200/9router

Quick start (use published image):

docker run -d \
  --name 9router \
  -p 20128:20128 \
  -v "$HOME/.9router:/app/data" \
  -e DATA_DIR=/app/data \
  vibecoder11200/9router:latest

→ Open http://localhost:20128

Build from source (dev):

git clone https://github.com/vibecoder11200/9router.git
cd 9router/app
docker build -t 9router .
docker run -d --name 9router -p 20128:20128 \
  -v "$HOME/.9router:/app/data" -e DATA_DIR=/app/data 9router

Container defaults:

PORT=20128
HOSTNAME=0.0.0.0

Useful commands:

docker logs -f 9router
docker restart 9router
docker stop 9router && docker rm 9router
docker pull vibecoder11200/9router:latest   # update to latest

Data persistence: $HOME/.9router/db/data.sqlite on host ↔ /app/data/db/data.sqlite in container.

Environment Variables

Variable	Default	Description
`JWT_SECRET`	Auto-generated (`~/.9router/jwt-secret`)	JWT signing secret for dashboard auth cookie (override to share across instances)
`INITIAL_PASSWORD`	`123456`	First login password when no saved hash exists
`DATA_DIR`	`~/.9router`	Main app data location (SQLite at `$DATA_DIR/db/data.sqlite`)
`PORT`	framework default	Service port (`20128` in examples)
`HOSTNAME`	framework default	Bind host (Docker defaults to `0.0.0.0`)
`NODE_ENV`	runtime default	Set `production` for deploy
`BASE_URL`	`http://localhost:20128`	Server-side internal base URL used by cloud sync jobs
`CLOUD_URL`	`https://9router.com`	Server-side cloud sync endpoint base URL
`NEXT_PUBLIC_BASE_URL`	`http://localhost:3000`	Backward-compatible/public base URL (prefer `BASE_URL` for server runtime)
`NEXT_PUBLIC_CLOUD_URL`	`https://9router.com`	Backward-compatible/public cloud URL (prefer `CLOUD_URL` for server runtime)
`API_KEY_SECRET`	`endpoint-proxy-api-key-secret`	HMAC secret for generated API keys
`MACHINE_ID_SALT`	`endpoint-proxy-salt`	Salt for stable machine ID hashing
`ENABLE_REQUEST_LOGS`	`false`	Enables request/response logs under `logs/`
`AUTH_COOKIE_SECURE`	`false`	Force `Secure` auth cookie (set `true` behind HTTPS reverse proxy)
`REQUIRE_API_KEY`	`false`	Enforce Bearer API key on `/v1/*` routes (recommended for internet-exposed deploys)
`HTTP_PROXY`, `HTTPS_PROXY`, `ALL_PROXY`, `NO_PROXY`	empty	Optional outbound proxy for upstream provider calls
`SEARXNG_URL`	`http://localhost:8888/search`	Endpoint for the built-in unauthenticated SearXNG web-search provider
`DS2API_URL` · fork	auto (loopback)	Override the DeepSeek Web sidecar URL
`DS2API_VERSION` · fork	`v4.6.2-rotation`	DS2API engine release tag (pulled from `vibecoder11200/ds2api`)
`DS2API_ADMIN_KEY` · fork	auto-generated	Override the DS2API sidecar admin secret
`XRAY_VERSION` · fork	`v26.3.27`	Xray-core binary release tag (auto-downloaded per OS/arch on first use)
`HEADROOM_URL`	`http://localhost:8787`	Headroom token-saver proxy endpoint

Notes:

Lowercase proxy variables are also supported: http_proxy, https_proxy, all_proxy, no_proxy.
.env is not baked into Docker image (.dockerignore); inject runtime config with --env-file or -e.
On Windows, APPDATA can be used for local storage path resolution.
INSTANCE_NAME appears in older docs/env templates, but is currently not used at runtime.

Runtime Files and Storage

Main app state: ${DATA_DIR}/db/data.sqlite (SQLite — providers, combos, aliases, keys, settings, usage history)
Auto backups: ${DATA_DIR}/db/backups/
Optional request/translator logs: <repo>/logs/... when ENABLE_REQUEST_LOGS=true
Both ${DATA_DIR} and ~/.9router resolve to the same location in a Docker container — the symlink /root/.9router -> /app/data is created at build time.

📊 Available Models

View all available models

Claude Code (cc/) - Pro/Max:

cc/claude-opus-5
cc/claude-sonnet-5
cc/claude-fable-5-1
cc/claude-fable-5
cc/claude-haiku-4-5-20251001

Codex (cx/) - Plus/Pro:

cx/gpt-5.6-sol
cx/gpt-5.5
cx/gpt-5.4
cx/gpt-5.4-mini
cx/gpt-5.3-codex-spark

GitHub Copilot (gh/):

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-5.3
glm/glm-5.1
glm/glm-5
glm/glm-4.7

MiniMax (minimax/) - $0.2/1M:

minimax/MiniMax-M3
minimax/MiniMax-M2.7
minimax/MiniMax-M2.5

Kimi (kimi/) - $9/mo flat:

kimi/kimi-k3
kimi/kimi-k2.7-code
kimi/kimi-k2.5
kimi/kimi-k2.5-thinking

Kiro (kr/) - Free (~50 credits/month, paid tiers above):

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

Grok CLI (gcli/) - OAuth (device-code):

gcli/grok-4.5, gcli/grok-4.5-high, gcli/grok-4.5-medium, gcli/grok-4.5-low

Perplexity Agent (perplexity-agent/) - API key, Responses API:

Cross-vendor routing: perplexity-agent/openai/gpt-5.5, perplexity-agent/anthropic/claude-sonnet-4-6, perplexity-agent/google/gemini-3.1-pro-preview, perplexity-agent/xai/grok-4.20-reasoning, plus Sonar. (Dynamic — fetched from /v1/models.)

Featherless (featherless/) - API key, OpenAI-compatible:

featherless/deepseek-v4-pro, featherless/glm-5.2, featherless/kimi-k2.7-code, and more.

Gemini Web (gemini-web/) · fork - cookie auth:

gemini-web/gemini-3-pro, gemini-web/gemini-3-flash, gemini-web/gemini-3-flash-thinking, gemini-web/gemini-3-flash-image, gemini-web/gemini-3-veo-video, gemini-web/gemini-3-audio (passthrough).

Genspark Web (genspark-web/) · fork - cookie auth:

genspark-web/gpt-5-pro, genspark-web/claude-sonnet-4-6, genspark-web/gemini-3-pro-preview, genspark-web/grok-4-0709 (append -search for web grounding), plus image models genspark-web/nano-banana-pro, genspark-web/fal-ai/flux-2 (passthrough).

DeepSeek Web (ds2api/) · fork - managed sidecar:

ds2api/<deepseek-models> — bare DeepSeek model names are auto-aliased on managed start.

🐛 Troubleshooting

"Language model did not provide messages"

Provider quota exhausted → Check dashboard quota tracker
Solution: Use combo fallback or switch to cheaper tier

Rate limiting

Subscription quota out → Fallback to GLM/MiniMax
Add combo: cc/claude-opus-5 → 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 (Kiro, OpenCode Free, Vertex) for non-critical tasks

Dashboard opens on wrong port

Set PORT=20128 and NEXT_PUBLIC_BASE_URL=http://localhost:20128

First login not working

Check INITIAL_PASSWORD in .env
If unset, fallback password is 123456

No request logs under logs/

Set ENABLE_REQUEST_LOGS=true

🛠️ Tech Stack

Runtime: Node.js 20+
Framework: Next.js 16
UI: React 19 + Tailwind CSS 4
Database: SQLite (better-sqlite3 / node:sqlite / sql.js fallback)
Streaming: Server-Sent Events (SSE)
Auth: OAuth 2.0 (PKCE) + JWT + API Keys

📝 API Reference

Chat Completions

POST http://localhost:20128/v1/chat/completions
Authorization: Bearer your-api-key
Content-Type: application/json

{
  "model": "cc/claude-opus-5",
  "messages": [
    {"role": "user", "content": "Write a function to..."}
  ],
  "stream": true
}

List Models

GET http://localhost:20128/v1/models
Authorization: Bearer your-api-key

→ Returns all models + combos in OpenAI format

📧 Support

Website: 9router.com
GitHub: github.com/vibecoder11200/9router
Issues: github.com/vibecoder11200/9router/issues

👥 Contributors

Thanks to all contributors who helped make 9Router better!

📊 Star Chart

🔀 Forks

This repository — vibecoder11200/9router: a feature-enhanced fork of upstream decolua/9router. Adds a managed V2Ray/Xray proxy (v2go), DeepSeek Web (DS2API) sidecar, rotating proxy pools/groups, Genspark & Gemini web-cookie providers, external tunnel URL, and a GitHub Releases distribution model. Track changes in CHANGELOG.md.

OmniRoute — A full-featured TypeScript fork of 9Router. Adds 36+ providers, 4-tier auto-fallback, multi-modal APIs (images, embeddings, audio, TTS), circuit breaker, semantic cache, LLM evaluations, and a polished dashboard. 368+ unit tests. Available via npm and Docker.

🙏 Acknowledgments

Built on the shoulders of giants:

CLIProxyAPI — original Go implementation that inspired this JavaScript port.
RTK — Rust token-saver. 9Router ports its compression pipeline to JS → −20-40% input tokens on every request.
Caveman by @JuliusBrussee — viral "why use many token when few token do trick". 9Router adapts its prompt → −65% output tokens.
Ponytail by @DietrichGebert — "lazy senior dev" skill. 9Router injects its YAGNI-first ladder → fewer tokens, less code, shorter diffs.

Huge thanks to these authors — without their work, 9Router's token-saving features wouldn't exist. ⭐ them on GitHub!

📄 License

MIT License - see LICENSE for details.

_{Built with ❤️ for developers who code 24/7}

Name		Name	Last commit message	Last commit date
.github		.github
.vscode		.vscode
cli		cli
docs		docs
gitbook		gitbook
i18n		i18n
images		images
open-sse		open-sse
plans		plans
public		public
scripts		scripts
skills		skills
src		src
tests		tests
.dockerignore		.dockerignore
.env.example		.env.example
.gitignore		.gitignore
.npmignore		.npmignore
AGENTS.md		AGENTS.md
CHANGELOG.md		CHANGELOG.md
CLAUDE.md		CLAUDE.md
DOCKER.md		DOCKER.md
Dockerfile		Dockerfile
LICENSE		LICENSE
README.md		README.md
README.zh-CN.md		README.zh-CN.md
captain-definition		captain-definition
custom-server.js		custom-server.js
docker-compose.yml		docker-compose.yml
eslint.config.mjs		eslint.config.mjs
https-server.js		https-server.js
install.ps1		install.ps1
install.sh		install.sh
jsconfig.json		jsconfig.json
next.config.mjs		next.config.mjs
package.json		package.json
postcss.config.mjs		postcss.config.mjs
start.sh		start.sh

Latest commit

History

Folders and files

Repository files navigation

9Router - FREE AI Router & Token Saver

🤔 Why 9Router?

🔄 How It Works

⭐ Fork Features

⚡ Quick Start

🔄 Update / Nâng cấp

Video Guides

🛠️ Supported CLI Tools

🌐 Supported Providers

🔐 OAuth Providers

🆓 Free Providers

🍪 Web-Cookie Providers · fork

🔑 API Key Providers (40+)

🏠 Self-hosted Providers

💡 Key Features

🚀 RTK Token Saver

🧠 Headroom Token Saver

🖼️ PXPipe Token Saver

🐴 Ponytail (Lazy Senior Dev)

🎯 Smart 3-Tier Fallback

📊 Real-Time Quota Tracking

🔄 Format Translation

👥 Multi-Account Support

🔄 Auto Token Refresh

🎨 Custom Combos

📝 Request Logging

💾 Cloud Sync

Cloud Runtime Notes

📊 Usage Analytics

🌐 Deploy Anywhere

💰 Cost & Strategy

Pricing at a glance

Billing reality

Example combos

❓ Frequently Asked Questions

📖 Setup Guide

🔌 Connect a provider

🆓 FREE providers (recommended)

Setup

Features

Engine / settings

Create a pool

Bind to a connection

How rotation behaves at runtime

Setup

Per-account proxies & rotating groups

Engine / env

Cursor IDE

Claude Code

Codex CLI

OpenClaw

Cline / Continue / RooCode

VPS Deployment

Docker

Environment Variables

Runtime Files and Storage

📊 Available Models

🐛 Troubleshooting

🛠️ Tech Stack

📝 API Reference

Chat Completions

List Models

📧 Support

👥 Contributors

📊 Star Chart

🔀 Forks

🙏 Acknowledgments

📄 License

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors