Hugging Face chat models now route through OpenRouter: 160 models from 38 labs
cpaik · x · 2026-11-17
OpenRouter founder Alex Atallah announced that chat models on the Hugging Face Hub now run through OpenRouter's routing. Open-weight models grew from 27% to 72% of OpenRouter tokens in under a year, driven by labs publishing weights on HF and dozens of competing providers.
Key points:
- 160 models from 38 labs, each request routed to the best provider by price, speed and uptime, with automatic failover
- Everything runs on the same Hugging Face token and API
- Users can send OpenRouter traces to their own HF Storage via Broadcast, ready for evals or fine-tuning
Related event: Hugging Face Partners with OpenRouter to Route Open-Model Inference(5 posts)→
More from Infra
- Baseten launches Voice Agent Stack; Realtime Voice hits 73.5% on τ²-Bench at ~$0.01/min — baseten · 2026-11-17
- Cloud vet: consulting firms steered clients into brittle single-vendor architectures — DavidLinthicum · 2026-11-17
- 300-run benchmark: Opus orchestrating Haiku costs 53% less and finds 72 errors vs 49 — BeginningSir7334 · 2026-11-17
- YC F26 Startup Maritime Launches a Managed Cloud for Running AI Agents — ycombinator · 2026-11-17
- Why Big Tech Fears Local AI: Security, Performance and Cost Advantages — DavidLinthicum · 2026-11-17
- Microsoft Details Maia Dataflow Architecture: 4719 Tflop/s fp8, Over 99% of Peak — thoefler · 2026-11-17