Leaderboard

Rank	Diff	Model	Score	95% CI	Votes	License
1	—	grok-4-fast xAI	1253		32	Proprietary
2	—	gemini-3-pro-preview Google	1170		28	Proprietary
3	—	deepseek-v3.1-terminus DeepSeek	1157		41	Open Source
4	—	gemini-3.1-pro-preview Google	1149		12	Proprietary
5	—	claude-sonnet-4.6 Anthropic	1131		21	Proprietary
6	—	claude-opus-4.1 Anthropic	1116		34	Proprietary
7	—	grok-4.1-fast xAI	1104		28	Proprietary
8	—	gpt-4.1 OpenAI	1095		35	Proprietary
9	—	qwen3-vl-235b-a22b-instruct Qwen	1072		24	Open Source
10	▲ 1	grok-4 xAI	1065		33	Proprietary
11	▼ 1	gpt-5.1 OpenAI	1060		26	Proprietary
12	—	kimi-k2-0905 Moonshot AI	1036		35	Open Source
13	—	gpt-5.2 OpenAI	1018		22	Proprietary
14	—	gpt-5-chat OpenAI	1009		35	Proprietary
15	—	o3 OpenAI	994		24	Proprietary
16	—	glm-4.6 Zhipu AI	991		26	Open Source
17	—	deepseek-r1-0528 DeepSeek	984		37	Open Source
18	—	deepseek-chat-v3.1 DeepSeek	974		23	Open Source
19	—	qwen3-235b-a22b-2507 Qwen	971		30	Open Source
20	—	qwen3-max Qwen	964		39	Proprietary
21	—	claude-opus-4 Anthropic	956		34	Proprietary
22	—	kimi-k2 Moonshot AI	922		28	Open Source
23	—	gemini-2.5-pro Google	918		31	Proprietary
24	—	claude-sonnet-4.5 Anthropic	916		37	Proprietary
25	—	gpt-4o OpenAI	916		12	Proprietary

Models are ranked by Bradley-Terry score on an Elo scale. 95% bootstrap confidence intervals show statistical uncertainty; once significance-aware ranking is enabled, models with overlapping intervals will share a tied rank.