From the source
Hugging Face has moved its Inference Providers service from its own router to OpenRouter, sending every language model request through OpenRouter so developers can run open-weight models with an existing Hugging Face token and the same API.
- OpenRouter picks the provider for each request based on live price, speed and reliability, and existing code stays the same.
- The router supports
OpenAI's Chat Completions and Responses APIs and
Anthropic's Messages API for every chat model, and Hugging Face offers setup guides for Claude Code, Codex, OpenCode, Pi and Hermes Agent.
- When a provider slows or fails, traffic moves away from it until it recovers, and a failed attempt is retried on the next best provider inside the same call; previously Inference Providers picked a single provider and returned its error.
- Users can still choose a routing policy or provider with suffixes such as :fastest, :cheapest or a provider name, and routing is cache-aware, keeping a conversation on the provider holding its prompt cache.
- Pricing does not change: users pay the provider's price on their Hugging Face bill with no markup, PRO users get $2 in monthly credits and Team and Enterprise organizations get $2 per seat.
- Custom provider API keys no longer work for language models, though they still work for image and video providers, and image, video and audio generation continue to run directly through Hugging Face Inference Providers.
- OpenRouter says its rebuilt routing cut capacity errors by 46% while improving time to first token by 16.9% and throughput by 16.3%, and Hugging Face calls this the first step of the partnership, with account linking and more tasks routed through OpenRouter coming next.