Sign in and sign up just work
Signing in from a new device, and creating a new account, now complete every step. Before this, some people got stuck partway through. If that was you, please try again.
The header also has a plain **Sign in** link. It opens the sign-in window on the page you are on, so you do not lose your place.
A models page that tells you what you can call
The models page now says what it counts. It shows how many models you can call today and how many more need your own provider key. One click shows them.
Every model now says how it is billed: from your **credits**, **free**, or with **your key**. The same label is in the API as pricing.billing, so your code can read it. See the models API.
The newest list leads with models you can call. The kind tabs wrap on phones instead of sliding sideways. The filters are simpler, with price, speed and release date in the same filter groups as the rest.
Bring your own key, with less clutter
The BYOK page is denser. You see your Pool status in one line, fewer divider lines, and more providers on a wide screen. Model pages that need your key say so in plain words and have one clear **Add your key** action.
New this week: perplexity/pplx-embed-v1-4b, an embedding model you can use with your own Respan key. Read how it works in the BYOK docs.
Fewer failed requests
If you ask for an anyrouter/ model name that does not exist, the request now goes to anyrouter/auto instead of failing. Add :free and it goes to the free route. Your original name stays in your request log. See routing.
Laguna XS 2.1 now moves to the next provider after 20 seconds of silence, not 36. When a provider hangs, your answer arrives sooner.
A cached answer now gets its own response id and time. Before, two requests could share one id, which made tracing hard. Caching is described in response caching.
Safer logs
Our server logs now hide credentials inside error messages too, such as a key that a provider repeats back in its error text. We already hid fields named like secrets. This closes the gap where a key appeared inside plain text.
Easier to learn from
Each use case with a film now plays it on the page. Nothing loads from the video site until you press play. The home page footer also lists our latest posts and news.
Related posts
MiniMax M3.1 Flash is free on every plan
MiniMax M3.1 Flash, a 1M-context coding and agent model, is free on AnyRouter for now. Any plan works, with no daily cap.
The top website for free LLM models — and how they're actually free
AnyRouter serves a real free LLM API across 170+ catalog models. Here's what people actually route to for free, and the three honest mechanisms that make it free: the anyrouter/free tier, the donated shared pool, and self-hosted on-device models.
Is there an unlimited free LLM API? An honest answer
Truly unlimited and truly free can't both be true forever — someone pays for the GPUs. But you can get very close: free tiers, your own provider quota at $0 markup, and a shared key pool. Here's how each one works.
Route your first request in 2 minutes
Start free with your own keys, or top up and pay per token. Get $4/mo in credits and free models on Go — $2/mo, or free when you donate a provider key.
Start free