9router
Open-source AI routing gateway that fronts 60+ providers behind one OpenAI-compatible endpoint at localhost:20128/v1.
If you pay for more than one AI coding subscription, 9Router is one of the few tools that turns that spend into throughput instead of idle quota. The 3-tier fallback is the core value, and RTK shipping on by default means you get the 20–40% tool_result compression without configuring anything. The free tier is real: Kiro AI, iFlow, Qwen and OpenCode Free mean $0/month routing if you accept their limits. Compare against OpenRouter — hosted, broader catalogue, but it can't pool your Claude Code or Copilot subscriptions. Think twice about the MITM bridge, which intercepts Antigravity, Copilot and Kiro IDE traffic against those vendors' policies.
Verified 4d ago · liveness 77/100 · cite: rightaichoice.com/tools/9router
- Developers running Claude Code, Cursor, Codex or Cline who keep hitting rate limits
- Anyone managing separate API keys and dashboards for every AI provider
- Power users pooling multiple AI subscriptions into one endpoint
- Hobbyists routing coding through free-forever providers at $0/month
- Non-technical users who don't want a CLI and npm install
- Buyers who need a vendor-backed uptime SLA
- Organizations requiring enterprise audit logging and compliance controls
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip 9Router if you hold only one AI subscription — the 3-tier fallback needs a second and third provider to do anything, and you'd be installing a CLI gateway to route to the same vendor you already pay.
GLM Coding and MiniMax bill you directly at $0.60 and $0.20 per million tokens — 9Router is free, but Tier 2 fallback capacity is not, so heavy fallback traffic shows up on those provider invoices.
9Router itself costs nothing to run — the Free Forever tier covers the router, all 60+ providers and the token savers. Your real spend is pass-through: Tier 2 providers bill you directly, with GLM Coding at $0.60 per million tokens, MiniMax at $0.20 per million and Kimi at $9/mo. That undercuts hosted gateway credit markups, and Tier 3's 8 free providers mean a hobbyist can route at $0/month.
In short
9router — Open-source AI routing gateway that fronts 60+ providers behind one OpenAI-compatible endpoint at localhost:20128/v1. Best for Developers running Claude Code, Cursor, Codex or Cline who keep hitting rate limits, Anyone managing separate API keys and dashboards for every AI provider, Power users pooling multiple AI subscriptions into one endpoint. Free to use.
What people actually say about 9router — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
24 mentions across 3 sources (Hacker News, YouTube, Lemmy) · researched Aug 27, 2026.
Average across the 3 sources that answered — each source counts once, not each post.
- +Free and open-source, no hidden fees
- +Supports 60+ providers with a unified API
- +3-tier fallback prevents rate-limit interruptions
- +Token optimization saves 20-65% on usage
- +Pools multiple subscriptions into one gateway
- −Setup requires command-line comfort and manual config
- −SQLite driver issues on Windows break default install
- −Risk of account bans from providers when pooling subscriptions
- −Documentation is sparse and scattered across tutorials
- −No official support; community is primary help
- • You must bring your own API keys or subscriptions
- • Running a VPS for 24/7 access has associated costs
- • Potential costs if you upgrade to paid tiers on providers
Viability Score
How well maintained and how widely used is 9router? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- OpenAI-compatible endpoint fronting 60+ AI providers at localhost:20128/v1
- Smart 3-tier fallback: subscriptions, then cheap paid providers, then free providers
- 9 service kinds: LLM chat, embeddings, TTS, STT, image, vision, video, web search, web fetch
- Smart Combos chain providers into one virtual provider with sticky round-robin
- Format translator between OpenAI, Anthropic and Gemini API formats
- Multi-account round-robin load balancing with automatic failover
- RTK token saver compresses tool_result input 20–40%, on by default
- Caveman Mode terse system prompt cuts output tokens up to 65%, five intensity levels
- Real-time quota tracking with reset countdown and cost estimation
- Cloudflare edge tunnel for remote access from 300+ global locations
- Proxy pools with Vercel relay and outbound IP rotation
- MITM bridge intercepts Antigravity, GitHub Copilot and AWS Kiro IDE traffic
- Per-tool setup guides for 10+ CLI and IDE clients
- Local visual dashboard for providers, keys, combos and usage
- Cloud sync for router configuration
About 9router
9Router is a free, open-source AI routing gateway you install locally with npm install -g 9router. It puts 60+ AI providers behind a single OpenAI-compatible endpoint at http://localhost:20128/v1, so Cursor, Claude Code, Codex, Cline, Kilo Code, Roo, Continue, Factory Droid, Hermes Agent and OpenClaw all point at one URL instead of a separate key and dashboard per vendor. The routing logic is a smart 3-tier fallback. Tier 1 spends the subscriptions you already hold (Claude Code, Codex, Gemini, Copilot). Tier 2 drops to cheap paid providers, including GLM Coding at $0.60 per million tokens, MiniMax at $0.20 per million and Kimi at $9/mo. Tier 3 lands on the 8 free-forever providers the vendor lists: Kiro AI, iFlow, Qwen, OpenCode Free, OpenRouter Free, NVIDIA NIM, Gemini and Cloudflare AI. Coverage spans nine service kinds rather than chat alone: LLM completions, embeddings, text-to-speech, speech-to-text, image generation, vision, video generation, web search and web fetch. Advanced pieces include Smart Combos that chain providers into one virtual provider with sticky round-robin, a format translator between OpenAI, Anthropic and Gemini, multi-account round-robin per provider, a Cloudflare edge tunnel, proxy pools with Vercel relay, and an MITM bridge that intercepts Antigravity, GitHub Copilot and Kiro IDE traffic. RTK compresses tool_result input by 20–40% and is on by default; Caveman Mode injects a terse system prompt and the vendor claims up to 65% output-token savings across five intensity levels. This suits developers who hit rate limits mid-task and already pay for more than one subscription. Each provider still bills you directly at its own rates, and the MITM bridge sits in a policy gray zone on the IDE tools it intercepts.
Behind the Verdict
9Router's pitch is narrow and honest: you already have AI capacity scattered across subscriptions, API keys and free tiers, and this routes it behind one endpoint. Strengths. The one-URL architecture (http://localhost:20128/v1) is genuinely simple — every OpenAI-compatible client works without an adapter. The 3-tier fallback is the differentiator: Tier 1 burns subscription quota you've already paid for, Tier 2 falls to GLM Coding at $0.60 per million tokens, MiniMax at $0.20 per million or Kimi at $9/mo, and Tier 3 lands on 8 free-forever providers. The two token savers are unusual bundling: RTK is lossless tool_result compression (git diff, grep, ls, find, tree) at 20–40%, on by default, and Caveman Mode trades verbosity for token count with five intensity settings. Real-time quota tracking with reset countdown and cost estimation means you see the fallback coming, not just its aftermath. Nine service kinds — embeddings, TTS, STT, image, vision, video, web search, web fetch alongside chat — go past what most coding-oriented routers expose. Format translation between OpenAI, Anthropic and Gemini means a Claude-format tool can reach a Gemini-only provider. Weaknesses. Setup is CLI-first: npm install -g 9router, then OAuth or API key per provider. The dashboard is local, at localhost:20128, so there's no hosted web app you can log into from a laptop you didn't set up. The MITM bridge that intercepts Antigravity, GitHub Copilot and AWS Kiro IDE traffic is the sharpest edge — the vendor's own copy says "mind each tool's policy," which is a fair warning rather than a feature note. Because 9Router itself doesn't sell tokens, your real bill is whatever GLM, MiniMax, Kimi and the rest charge you directly, so cost control is still a per-provider discipline. And the value proposition collapses if you hold only one subscription. Where it fits. Solo developers and small teams running Claude Code, Cursor or Codex daily who keep hitting rate limits mid-task. Hobbyists who want to code on free-forever providers at $0/month with RTK trimming the input. Anyone tired of five dashboards for five keys. Where it doesn't. Non-technical buyers who want nothing to install. Teams that need vendor-backed uptime commitments. Organizations that require compliance controls over interception of IDE subscription traffic — the MITM bridge is exactly the kind of thing a security review flags.
Researching 9router? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas 9router actually fits — and what changes day-one when you adopt it.
Installs via npm install -g 9router, connects Claude Code by OAuth and adds GLM Coding plus iFlow as Tier 2 and Tier 3, then points Claude Code at http://localhost:20128/v1.
Outcome: Work continues past the Claude Code quota wall instead of stalling until reset, with the dashboard showing which tier handled each request.
Connects Claude Code, OpenAI Codex, Gemini and Copilot as Tier 1, adds multiple accounts per provider for round-robin, and monitors live quota and reset countdowns from the local dashboard.
Outcome: Quota that used to expire unused each month gets spent across whichever subscription still has headroom.
Skips paid providers entirely and connects the free-forever set — Kiro AI, iFlow, Qwen, OpenCode Free — with RTK left on by default.
Outcome: Coding assistance runs without API credits, and RTK's 20–40% tool_result compression stretches the free quotas further.
Use Cases
- Route Claude Code, Codex or Cursor through one endpoint that falls back automatically when you hit a rate limit mid-task.
- Run Copilot or Antigravity in your IDE through the MITM bridge and reach any provider you've connected to 9Router.
- Code on the 8 free-forever providers — Kiro AI, iFlow, Qwen, OpenCode Free — at $0/month with no API credits.
- Chain GLM Coding at $0.60 per million tokens, MiniMax at $0.20 per million and Kimi at $9/mo as Tier 2 fallbacks instead of burning subscription quota.
- Turn on RTK to compress tool_result content from git diff, grep, ls, find and tree by 20–40% before it reaches the model.
- Use Caveman Mode at one of five intensity levels to trim verbose model output on long agent runs.
- Expose your local router through the Cloudflare edge tunnel to use it from any machine in 300+ locations.
- Rotate outbound IPs via Vercel relay proxy pools to reach region-restricted providers.
Models Under the Hood
as of 2026-09-24
Limitations
- Setup is CLI-based: npm install -g 9router, then OAuth or an API key per provider, with a local dashboard at localhost:20128 rather than a hosted web app.
- The value depends on you already holding multiple AI subscriptions or API keys — with a single provider the 3-tier fallback has nothing to fall back to.
- The MITM bridge intercepts Antigravity, GitHub Copilot and AWS Kiro IDE traffic, and the vendor's own guidance is to mind each tool's policy, which puts that configuration in a gray zone.
- Provider costs are billed by each provider directly, not by 9Router, so the built-in savings figures (RTK 20–40%, Caveman up to 65%) are the vendor's own measurements and vary with your workload.
- No dedicated mobile app or native hosted web client is described in the evidence.
as of 2026-10-04
Verification history
We have re-verified 9router 8 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 8 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Where the pricing makes sense
The company stage and team size where 9router's pricing actually pencils out — and where peers do it cheaper.
9Router itself costs nothing to run — the Free Forever tier covers the router, all 60+ providers and the token savers. Your real spend is pass-through: Tier 2 providers bill you directly, with GLM Coding at $0.60 per million tokens, MiniMax at $0.20 per million and Kimi at $9/mo. That undercuts hosted gateway credit markups, and Tier 3's 8 free providers mean a hobbyist can route at $0/month.
Setup time & first value
How long it actually takes to get something useful out of 9router — broken out by persona, not the marketing-page minute.
Roughly 5 minutes for the single-provider path: npm install -g 9router, run 9router, open the dashboard at localhost:20128. Budget 20–40 minutes for a full 3-tier setup, since each provider needs its own OAuth or API key and Smart Combos add a configuration step. The MITM bridge for Copilot, Antigravity or Kiro IDE traffic is the slowest piece and needs policy review first.
Switching to or from 9router
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From a single-provider client (Claude Code, Codex, Cursor): keep the subscription, add it as Tier 1, and repoint the client's base URL to http://localhost:20128/v1.
- →From OpenRouter: add your OpenRouter key as one provider, so existing credits sit alongside your subscriptions and free tiers.
- →From manually juggling API keys per tool: import each key once into the dashboard and let the 3-tier fallback pick the provider.
- ↗To direct provider endpoints: restore each client's base URL to the vendor's own API and re-enter that vendor's key.
- ↗To a hosted gateway: export your provider keys from the local dashboard, then re-add them at the hosted service.
- ↗To nothing: 9Router is local and open-source, so uninstalling the global npm package ends the setup.
Integrations
Resources & Guides
Tutorials & Learning
YouTube returned 6 videos for “9router”, and we withheld 6: 6 could not be judged, because “9router” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about 9router.
Official links
Tools that pair well with 9router
Common stack mates teams adopt alongside 9router, with the specific reason each pairing earns its keep.
OmniRoute
Self-hosted open-source AI gateway that routes 352 LLM providers through one OpenAI-compatible endpoint with automatic fallback.
LiteLLM
Self-hosted AI gateway putting 140+ providers, MCP servers, and agents behind one OpenAI-compatible API.
Agnes AI
Free multimodal API gateway from Singapore's Sapiens AI with in-house text, image, video and audio models behind OpenAI-compatible endpoints
Featured Head-to-Head Comparisons
9router vs Spider Cloud
Choose Spider Cloud if you need high-throughput, structured web data extraction for AI agents or RAG pipelines. Choose 9Router if you're a developer tired of juggling multiple AI subscriptions and want a unified endpoint with auto-fallback for coding assistance. They solve different problems—one feeds AI models with fresh data, the other optimizes access to those models.
9router vs Voyage Ai
For enterprise RAG needing high-accuracy retrieval on specialized documents, Voyage AI's domain-specific models and low-dimensional embeddings are unmatched. For developers tired of API rate limits and juggling multiple AI subscriptions, 9Router's free, open-source unified endpoint with auto-fallback across 60+ providers is a game-changer—but recent news highlights potential security fingerprinting risks. Choose based on your core need: retrieval accuracy vs. endpoint availability.
9router vs Temporal Ai
Temporal AI and 9router serve entirely different needs: Temporal is for building fault-tolerant AI workflows that survive crashes (trusted by OpenAI, Replit), while 9router is a free local router to unify 60+ providers and never hit rate limits. Choose Temporal if you need durable execution and state recovery; choose 9router if you're a developer tired of API caps and want a single endpoint with auto-fallback.
Alternatives to 9router
View all
Frequently Asked Questions
Categories
Topics
Used 9router? Help shape our editorial sentiment research.