Run any model on expert-written tests. Every score checkable.

Pick a test set from the community, or write one for your field.

Runs
36Runs
Models tested
9Models tested
Public test sets
22Public test sets
Questions in them
576Questions in them

Featured test sets

Browse all test sets

The open-model market

250 open · 191 closed · 57 labs · data from Hugging Face, LMArena, OpenRouter

Open-weight models

Score by size

Plot against
Best for its size
Gap
Coding
−55
Math
−41
Expert
−64
Chinese
−96
140015001600

Arena scores: LMArena, CC BY 4.0, 2026-11-11. Measured on served versions.

Discussed, last 30 days

On BenchClub