Benchmarks

⌘K↑	Publisher↕	Source rows↕	Mapped↕	Published↕	Unmapped↕	Excluded↕	Catalog coverage↕	Status↕
Aider Polyglotpolyglot-225-5dc9490bb35f	Aidercoding	69	13	10	56	3	4.5%	Mapped rows excluded56 source IDs need review
ARC-AGI-1verified-v1-613c385b0c93	ARC Prize Foundationabstract-reasoning	265	224	48	41	176	21.7%	Mapped rows excluded41 source IDs need review
ARC-AGI-2verified-v2-be40992860f9	ARC Prize Foundationabstract-reasoning	265	220	48	45	172	21.7%	Mapped rows excluded45 source IDs need review
Berkeley Function Calling Leaderboardv4-ede5081a24bc	Berkeley Function Calling Leaderboardtool-use	109	28	18	81	10	8.1%	Mapped rows excluded81 source IDs need review
LiveBench2026-06-25-90f8d44da965	LiveBenchgeneral-language	71	58	55	13	3	24.9%	Mapped rows excluded13 source IDs need review
LMArena Agent Arenaagent-2026-10-16-a64ca3b12b6e	LMArenaagentic	52	18	18	34	0	8.1%	Mapping incomplete34 source IDs need review
LMArena Document Arenadocument-2026-09-13-a64ca3b12b6e	LMArenaknowledge-work	44	39	33	5	6	14.9%	Mapped rows excluded5 source IDs need review
LMArena Search Arenasearch-2026-10-13-a64ca3b12b6e	LMArenasearch	42	31	30	11	1	13.6%	Mapped rows excluded11 source IDs need review
LMArena Text Arenatext-2026-10-15-a64ca3b12b6e	LMArenageneral-language	416	95	86	321	9	38.9%	Mapped rows excluded321 source IDs need review
LMArena Vision Arenavision-2026-10-09-a64ca3b12b6e	LMArenamultimodal	165	67	58	98	9	26.2%	Mapped rows excluded98 source IDs need review
ToneBench2026-10-17-10-task-e9c76358afb2	Towards AIwriting	134	93	55	41	38	24.9%	Mapped rows excluded41 source IDs need review

Select a benchmark to view its model leaderboard. Mapped rows have an exact catalog identity; unmapped source IDs remain quarantined.