Skip to content

PlatformsHugging Face

Hugging Face Inference Providers switches language-model inference routing to OpenRouter

Hugging Face Inference Providers at a glance

Hugging Face Inference Providers now routes language-model requests through OpenRouter across 80+ providers, with pricing unchanged and no markup. The inference routing change is live, but custom provider API keys no longer work for language models — a breaking change for existing setups.

Ask your AI about this story

Ask what changed in Hugging Face Inference Providers and OpenRouter and what to check before using it in your business. Weigh pricing, what it connects to, setup effort and how hard it would be to leave later.

Sources checked Primary source (opens in a new tab)

What changed with Hugging Face Inference Providers

Hugging Face announced on November 16, 2026 that Hugging Face Inference Providers has switched its inference routing for language models from its own custom router to OpenRouter's, so requests are routed across more than 80 providers based on live price, speed and reliability, and retried on another provider if the first one fails. User code stays the same, and the router supports the OpenAI Chat Completions and Responses APIs and Anthropic's Messages API, giving apps one integration for open-weight models and proprietary models alike.

Pricing is unchanged: you pay the provider price on your Hugging Face bill with no markup, PRO includes $2 of credits a month, and Team and Enterprise include $2 per seat. One breaking change matters: custom provider API keys no longer work for language models, though they still work for image and video providers. OpenRouter says it offers 600+ models, serves more than 800 trillion tokens a month to over 10 million users, and that the open-weight share of its tokens rose from 26.8% in October 2025 to 71.7% at the end of September 2026; by OpenRouter's measurements, its new routing cut capacity errors by 46% and improved time to first token by 16.9% and throughput by 16.3%.

Why Hugging Face Inference Providers matters for businesses

Why it matters: a business building on Hugging Face can now reach hundreds of models through one account and one bill, with requests automatically moved to a working provider when one is slow or down — less integration work and fewer outages to manage. Teams using their own provider keys for language models must switch to the new routing before their apps break, while the unchanged no-markup pricing keeps the cost of testing open-weight models against proprietary ones predictable.

Put it to work: AI services

Get AI News by email

Verified AI stories, each with its official source, on Monday and Thursday mornings. Add breaking news to hear about the biggest ones as soon as we publish them.

Every story links its official source

Free, in English or Spanish

Unsubscribe with one click, any time

What to send

Sources for this story

  1. Official

Last verified: . IA Compa summarizes public announcements for business readers. We are not affiliated with the companies mentioned. Always confirm details, pricing, and availability with the official source.

  • #hugging-face
  • #openrouter
  • #hugging-face-inference-providers
  • #inference-routing
  • #open-weight-models
  • #ai-platforms

Want this working in your business?

We turn new AI capabilities into agents, voice receptionists, and automations you can actually use.