vals Mmlu Pro

125 models evaluated

#	Model	Provider	Type	Score
1	Anthropic	anthropic/claude-opus-5	Claude Opus 5	91.59%
2	Anthropic	anthropic/claude-fable-5	Claude Fable 5	91.50%
3	Google	google/gemini-3.1-pro-preview	Gemini 3.1 Pro Preview	90.99%
4	Gemini 3 Pro Preview	Google	90.10%
5	Claude Opus 4.7	Anthropic	89.87%
6	Claude Opus 4.8	Anthropic	89.58%
7	Gemini 3.5 Flash	Google	89.52%
8	Qwen3.7 Max	Alibaba	89.31%
9	Gemini 3.6 Flash	Google	89.28%
10	Grok 4.5	xAI	89.22%
11	Claude Opus 4.6 Thinking	Anthropic	89.11%
12	GPT-5.6 Sol	OpenAI	89.10%
13	Muse Spark 1.1	Meta	88.73%
14	Gemini 3 Flash Preview	Google	88.59%
15	GPT-5.5	OpenAI	88.14%
16	Kimi K3	Moonshot AI	87.97%
17	Claude Opus 4.1 20250805 Thinking	Anthropic	87.92%
18	Qwen3.6 Plus	Alibaba	87.67%
19	Kimi K2.6	Moonshot AI	87.57%
20	Claude Sonnet 5	Anthropic	87.55%
21	GPT-5.4	OpenAI	87.48%
22	Claude Sonnet 4.5 20250929 Thinking	Anthropic	87.36%
23	Claude Sonnet 4.6	Anthropic	87.34%
24	Muse Spark	Meta	87.32%
25	Claude Opus 4.5 20251101 Thinking	Anthropic	87.26%
26	DeepSeek V4 Pro	DeepSeek	87.25%
27	Claude Opus 4.1	Anthropic	87.21%
28	Qwen3.5 Plus Thinking	Alibaba	87.18%
29	MiniMax M2.1	MiniMax	87.05%
30	GLM 5.1	Zhipu AI	86.90%
31	GLM 5.2	Zhipu AI	86.71%
32	GPT-5.6 Terra	OpenAI	86.66%
33	GPT-5	OpenAI	86.54%
34	GPT-5.1	OpenAI	86.38%
35	Inkling	Thinkingmachines	86.30%
36	Grok 4.20 0309 Reasoning	xAI	86.25%
37	Gemini 3.1 Flash Lite Preview	Google	86.24%
38	GPT-5.2	OpenAI	86.23%
39	Claude Opus 4	Anthropic	86.17%
40	GPT-5.6 Luna	OpenAI	86.04%
41	GLM 5 Thinking	Zhipu AI	86.03%
42	Kimi K2.5 Thinking	Moonshot AI	85.91%
43	Grok 4.3	xAI	85.84%
44	Gemini 3.5 Flash Lite	Google	85.84%
45	Nemotron 3 Ultra 550b A55b	Nvidia	85.76%
46	O3	OpenAI	85.59%
47	Claude Opus 4.5	Anthropic	85.59%
48	Grok 4 0709	xAI	85.30%
49	Qwen3 Max	Alibaba	84.98%
50	DeepSeek V3p2 Thinking	Fireworks AI	84.92%
51	Mimo V2.5 Pro	Xiaomi	84.59%
52	GPT-5.4 Mini	OpenAI	84.55%
53	Qwen3 Max	Alibaba	84.36%
54	MiniMax M3	MiniMax	84.22%
55	Grok 4.1 Fast Reasoning	xAI	84.18%
56	Qwen3.5 Flash	Alibaba	84.06%
57	Gemini 2.5 Pro Exp 03 25	Google	84.06%
58	Claude Sonnet 4 20250514 Thinking	Anthropic	83.86%
59	Gemini 2.5 Flash Preview 09 2025	Google	83.69%
60	Gemini 2.5 Flash Preview 09 2025 Thinking	Google	83.66%
61	Qwen3 Max Preview	Alibaba	83.54%
62	O1	OpenAI	83.49%
63	DeepSeek R1	Fireworks AI	83.18%
64	DeepSeek V3p2	Fireworks AI	83.06%
65	Mimo V2.5	Xiaomi	82.93%
66	GLM 4.7	Zhipu AI	82.74%
67	Claude 3.7 Sonnet 20250219 Thinking	Anthropic	82.73%
68	GPT-5 Mini	OpenAI	82.23%
69	GLM 4.6	Zhipu AI	82.20%
70	Grok 3 Mini Fast High Reasoning	xAI	81.37%
71	Qwen3 235b A22b	Fireworks AI	81.25%
72	GLM 4.5	Zhipu AI	81.22%
73	Kimi K2 Thinking	Moonshot AI	81.07%
74	Claude 3.7 Sonnet	Anthropic	80.66%
75	O4 Mini	OpenAI	80.56%
76	GPT-4.1	OpenAI	80.50%
77	MiniMax M2.7	MiniMax	80.43%
78	MiniMax M2.5	MiniMax	80.09%
79	Grok 3 Mini Fast Low Reasoning	xAI	80.01%
80	Grok 3	xAI	79.95%
81	Mistral Large 2512	Mistral AI	79.82%
82	Grok 4 Fast Reasoning	xAI	79.70%
83	DeepSeek V3 0324	Fireworks AI	79.47%
84	Claude Sonnet 4	Anthropic	79.43%
85	Llama4 Maverick Instruct Basic	Fireworks AI	79.42%
86	Moonshotai Kimi K2 Instruct	Together AI	79.39%
87	GPT Oss 120b	Fireworks AI	79.17%
88	Gemini 2.5 Flash Lite Preview 09 2025 Thinking	Google	79.12%
89	Claude Haiku 4.5 20251001 Thinking	Anthropic	78.72%
90	O3 Mini	OpenAI	78.69%
91	Gemini 2.5 Flash Lite Preview 09 2025	Google	78.64%
92	Claude 3.5 Sonnet	Anthropic	78.40%
93	Gemini 2.0 Flash 001	Google	77.38%
94	GPT-4.1 Mini	OpenAI	77.22%
95	GPT-5.4 Nano	OpenAI	77.17%
96	GPT-5 Nano	OpenAI	76.07%
97	Grok 2 1212	xAI	75.47%
98	Mistral Medium 3.5	Mistral AI	75.33%
99	Gemini 1.5 Pro 002	Google	75.29%
100	Mistral Medium 2505	Mistral AI	75.29%
101	Grok 4.1 Fast Non Reasoning	xAI	75.21%
102	GPT-4o	OpenAI	74.13%
103	DeepSeek V3	Fireworks AI	73.82%
104	GPT-4o	OpenAI	72.56%
105	GPT Oss 20b	Fireworks AI	71.64%
106	Langston Nim Nvidia Llama 3.3 Nemotron Super 49b V1 42e84561	Together AI	70.78%
107	Grok 4 Fast Non Reasoning	xAI	70.34%
108	Meta Llama Llama 3.3 70B Instruct Turbo	Together AI	69.86%
109	Mistral Large 2411	Mistral AI	69.71%
110	Meta Llama Llama 4 Scout 17B 16E Instruct	Together AI	69.63%
111	Langston Nim Nvidia Llama 3.3 Nemotron Super 49b V1 42e84561 Thinking	Together AI	69.58%
112	Laguna Xs.2	Poolside	69.41%
113	Command A 03 2025	Cohere	69.17%
114	Magistral Medium 2509	Mistral AI	68.66%
115	Mistral Small 2503	Mistral AI	66.02%
116	Gemini 1.5 Flash 002	Google	65.61%
117	Mistral Small 2402	Mistral AI	64.44%
118	Claude 3.5 Haiku	Anthropic	64.12%
119	GPT-4.1 Nano	OpenAI	63.48%
120	Laguna M.1	Poolside	63.48%
121	GPT-4o Mini	OpenAI	62.73%
122	Magistral Small 2509	Mistral AI	62.13%
123	Jamba Large 1.6	AI21 Labs	49.78%
124	Command R Plus	Cohere	44.00%
125	Jamba Mini 1.6	AI21 Labs	30.28%