Introducing AnyRouter π
AnyRouter is for building your next AI app. 200+ models, one key.
- Start at $0, donate an idle key to the shared pool
- Get millions of free tokens every month
- anyr CLI that wraps your coding agent
- One key for 200+ models, so you can swap models without changing code
- Many more
β https://t.co/taIV53PGDT
Your apps can reach local models from your computer through AnyRouter Relay.
$ anyr relay start
Your Mac dials out over a secure tunnel, so no ports are open. Apple Foundation Model, Perplexity's Decider, Ollama or LM Studio answers, and only your own requests reach your Mac.
Shipped this week on @AnyRouterHQ:
- Gemini 4 Hydrogen, Inkling 2 Small and Qwen4 Max
- Alibaba Cloud and QwenCloud Token Plan keys as BYOK providers
- A Live data panel on artifact pages
- Cheap models ranked by score per dollar
Artifact pages that read live data now show a Live data icon in the header: each source's host, how often it refreshes and whose data it is. View data mapping shows which widgets read what.
https://t.co/MiSkstT8Q2
Your key stays yours. AnyRouter adds no fee on bring-your-own-key requests, and Google's limits and terms apply to your project.
Add a key:
https://t.co/AlvgdrPjyK
Google AI Studio is a bring-your-own-key provider on @AnyRouterHQ.
Add your key once, then call your Gemini models through the same OpenAI-compatible endpoint you use for everything else.
GPT-6.2 Sol and Astra are 95% off on @AnyRouterHQ, for a limited time.
- Sol: $0.10 input and $0.50 output per 1M tokens, down from $2 and $10.
- Astra: $0.40 input and $2 output, down from $8 and $40.
GPT-6.2 Sol and Astra are 95% off on @AnyRouterHQ, for a limited time.
- Sol: $0.10 input and $0.50 output per 1M tokens, down from $2 and $10.
- Astra: $0.40 input and $2 output, down from $8 and $40.