Never taxed
Bring your own keys, keep every cent.
Route through your own provider keys and AnyRouter takes no cut — no deposit fee, no per-token markup, no cap. What you pay upstream is exactly what you pay.
Every SDK.
One gateway.
One OpenAI-compatible API, CLI and MCP gateway in front of every provider — and every other router. Free models, $4 credit every month on Go, and your own keys at no markup.
Free to start — $4 credit (~16M tokens) every month on Go, no card if you donate a key. Unlock the shared pool for $2/mo or by donating one working key.
import os
from openai import OpenAI
client = OpenAI(
base_url="https://anyrouter.dev/api/v1",
api_key=os.environ["ANYROUTER_API_KEY"],
)
resp = client.chat.completions.create(
model="anyrouter/auto",
messages=[{"role": "user", "content": "Hi"}],
)200 OKone line changed · base_url
$ anyr claudeShared pool
Crowdsourced free tokens
We pool everyone's free & trial keys into one big shared pool — free tokens for everyone. Donate a key, grow the pool, ride free.
Latest releaseOct 4
Mistral Small 4Mistral's Apache 2.0 hybrid model, which folds its instruct, reasoning and coding lines into one. 256k context, image input, $0.15 / $0.60 per 1M. BYOK.
Deals
Ongoing promotions
Free models and price cuts across the catalog — 26 live now (26 free).
- Moonshot AI
Kimi K3
Free- Input / 1M
$3.00Free- Output / 1M
$15.00Free
- thinkingmachines
Inkling
Free- Input / 1M
$1.00Free- Output / 1M
$4.05Free
- nvidia
Nemotron 3 Ultra 550B
Free- Input / 1M
$0.60Free- Output / 1M
$3.60Free
01
One endpoint. Every model.
Keep your SDK. One upstream rate-limits — your request doesn't notice.
Point your SDK at one base URL and reference models as provider/model. Every call then runs the same pipeline: authenticate, pick an upstream, retry the next one on failure, cache what repeats, and log every attempt.
YOUR CLIENTS
SDK
CLI
coding agent
app
AnyRouter
gateway
Auth + keys
Cache
Usage + logs
Router
Router
MCP servers
GitHub
Context7
Notion
Firecrawl
Linear
Sentry
Routerresolves to
presets
anyrouter/free
anyrouter/agent
@preset/my-preset
219+ MODELS · 17 PROVIDERS
deepseek/*
qwen/*
openai/*
anthropic/*
google/*
mistral/*
Between your call and the model · per request
Key balancing
Spreads load across your keys and quarantines burned ones automatically — no client changes.
Automatic failover
Retries and falls back across providers and pooled keys the instant one errors or rate-limits.
Prompt caching
Reuses cached context across calls to cut repeat token cost and time-to-first-token.
Per-request logs
Every fallback hop, token count and cost captured per attempt — traced and queryable.
03
Everything in one gateway
One key. Every capability you'd otherwise wire up yourself.
No add-ons, no tiers gating the basics. Every capability below ships on the same base URL and the same API key.
API & compatibility
Routing & reliability
Models & catalog
Keys & cost
Agents & tooling
04
Why teams switch
Three things that are structurally hard to copy.
Observable
Debug any request.
Every fallback hop, status, latency, token count and cost — captured per attempt. A “debug this request” trace nobody else ships at this tier.
Portable
Your config follows you.
Keys, presets and skills live in AnyRouter, inject locally on demand, and wipe clean on exit. Same setup on your laptop, a server, CI, or a teammate's box.
06
Unlock Go — two doors, same room
How to join
AnyRouter runs on Go — $2/mo, or donate one free-tier provider key (donors ride free). Every Go comes with $4 credit each month.
- #1No card
Create your account
Sign up in seconds — no card required. You land on the dashboard with your first API key ready to copy.
- #2Two doors
Pick your door into Go
Go is the plan that switches AnyRouter on — two doors, same room. Walk through whichever suits you:
- #3Every month
You're in
Go comes with $4 credit every month (~16M tokens on fast models), 1000 free-model requests a day, and access to the shared key pool.
07
The cheapest way in
Three ways to start
A $4 monthly credit and free models come with Go — $2/mo, or free when you donate a provider key. Or bring your own keys at no markup, no card required.
$4credit / month
How far $4 goes on fast models
- GPT-5 Nano29.1M
- GLM-4.7 Flash27.6M
- DeepSeek-V4-Flash75.4M
- Gemini 2.5 Flash-Lite22.9M
- Claude Haiku 4.52.0M
- Gemini 3.5 Flash1.2M
- Free models1000 req/day
Tokens for $4, blended 75% input / 25% output.
1000req / day · $0
- Apodex 1.1 Mini
- Apple Foundation Model (on-device)
- DeepSeek-V4-Flash
- Dots3-Note Preview
- Gemini 3 Flash
- Gemma 3n E2B (on-device)
- Gemma 4 31B
- Ling-3.0-flash-Sante
- Ling-3.1-flash
- LFM 2.5 2.6B
- Nemotron-3 Super 120B
- Nemotron 3 Ultra 550B
$0markup
14 providers ship a free tier — billed by the provider only, never by us.
- Freebuff+ Add keySee how
- Ramp Router+ Add keySee how
- OpenCode Zen+ Add keySee how
- Cerebras+ Add keySee how
- Google AI Studio+ Add keySee how
- Groq+ Add keySee how
- Mistral+ Add keySee how
- NVIDIA NIM+ Add keySee how
- Ollama Cloud+ Add keySee how
- OpenRouter+ Add keySee how
- QwenCloud+ Add keySee how
- AIHubMix+ Add keySee how
- ZenMux+ Add keySee how
- Hugging Face+ Add keySee how
When you need more
Full pricingGo
- $4 in credits every month (~16M tokens on fast models)
- Free-tier models — 1000 requests/day
- Shared key pool access
- 1 workspace · 7-day logs · 60 req/min
Pro
$10
per month$11.05 charged
- $20 in credits every month (~84M tokens on fast models)
- Unlimited usage — no 5-hour window limits
- Unlimited cloud skills & prompts + private bundles
- 3 workspaces · 30-day logs · 1200 req/min
Pro+
Best value$45
per month$47.89 charged
- $100 in credits every month (~420M tokens on fast models)
- 1800 req/min — our highest rate ceiling
Max
$100
per month$105.79 charged
- $300 in credits every month (~1.26B tokens on fast models)
- 1800 req/min — our highest rate ceiling
08
Sign in with AnyRouter
Add AI to your app — your users bring their own AnyRouter
One OAuth button hands your app a scoped, temporary key per user. Every request is billed to that user's own account — no API keys for you to collect, store, or pay for.
Your app's sign-in screen
Click it — see what your users see.
AnyRouter · Authorize
Acme AI
acme.app wants to sign you in
This app will be able to
- Send inference requestsRun AI billed to your account
- Read profileYour name and avatar
It can't manage your keys, billing, or settings. Access expires automatically and can be revoked any time.
- 01
OAuth 2.1 + PKCE
Standard authorization-code flow, public client, no secret to leak. Works from a browser-only app.
- 02
Inference-only scoped key
The token can run AI requests and read a basic profile — never keys, billing, or account settings.
- 03
Per-user billing & revocation
Each user pays from their own credits or free tier, and can revoke your app any time from their dashboard.
09
219+ models · 17 providers · October 4, 2026
One growing catalog, always current
Mistral's Apache 2.0 hybrid model, which folds its instruct, reasoning and coding lines into one. 256k context, image input, $0.15 / $0.60 per 1M. BYOK.
10
FAQ
Frequently asked questions
What is AnyRouter?
One OpenAI-compatible API, CLI and MCP gateway in front of every provider — and every other router. Free models, $4 credit every month on Go, and your own keys at no markup.
How do I call models through AnyRouter?
Point your SDK at one base URL and reference models as provider/model. Every call then runs the same pipeline: authenticate, pick an upstream, retry the next one on failure, cache what repeats, and log every attempt.
Do you take a cut if I bring my own keys?
Route through your own provider keys and AnyRouter takes no cut — no deposit fee, no per-token markup, no cap. What you pay upstream is exactly what you pay.
How do I get started?
Sign up in seconds — no card required. Go is the plan that switches AnyRouter on: pay $2 / month, or donate 1 free key. Go comes with $4 credit every month, unlimited free models, and access to the shared key pool.