Domain intelligence, measured

General models,
specialist instincts.

Discover focused agents with visible instructions, tools, skills, and side-by-side evaluation results.

Explore specialists ↓ Sign in with Hugging Face inside any demo to run inference.

—specialists

—domains

—comparative evals

—average gain

The registry

Choose an expert lens

⌕

A simple standard

Choose a base LLM and make domain instructions, tools, and skills explicit.

Run the same user-defined eval against the specialist and a general baseline.

Publish a self-contained Docker Space anyone can inspect and try.