2026-09-10 AI大模型排行

数据来源：BenchLM，共 228 个上榜模型，按 BenchAlign 综合分排序。分类分取自公开分类榜，未收录的结果显示为「—」。评分方法

排行	模型名	综合分	智能体	编程	推理	多模态	知识	多语言	指令遵循	数学	组织	开放类型

	Claude Fable 5.1	84.61	80.16	84.28	79.40	—	86.68	—	—	—	Anthropic	Proprietary
	GPT-6 Astra	84.08	70.38	74.57	89.50	—	82.16	—	—	—	OpenAI	Proprietary
	Claude Opus 5	81.97	78.07	75.84	75.70	88.70	82.28	—	—	—	Anthropic	Proprietary
4	Claude Fable 5	81.43	74.57	76.99	77.60	—	83.29	—	78.50	—	Anthropic	Proprietary
5	GPT-5.6 Sol	80.63	70.05	74.47	70.40	87.50	80.48	—	89.10	—	OpenAI	Proprietary
6	Gemini 3.8 Flash	75.61	66.31	66.64	76.90	82.80	74.91	—	—	—	Google	Proprietary
7	Kimi K3	74.82	71.91	67.95	78.50	89.50	72.00	—	—	—	Moonshot AI	Pending
8	Claude Opus 4.8	72.22	62.47	66.41	55.50	87.80	71.41	—	75.40	65.40	Anthropic	Proprietary
9	GPT-5.5	72.07	61.04	67.60	64.00	71.30	72.85	—	93.20	—	OpenAI	Proprietary
10	Qwen3.8 Max	71.63	67.16	60.80	—	87.40	68.75	—	90.70	—	Alibaba	Open Weight
11	GPT-5.6 Terra	71.13	60.20	67.01	63.70	77.10	69.16	—	87.10	—	OpenAI	Proprietary
12	GPT-5.4	70.85	52.53	53.85	57.20	69.30	69.21	—	90.60	—	OpenAI	Proprietary
13	Muse Spark 1.2	70.51	61.12	59.99	—	—	70.76	—	—	—	Meta	Proprietary
14	Muse Spark 1.1	70.41	59.39	59.68	—	—	70.66	—	—	—	Meta	Proprietary
15	Claude Opus 4.7	70.29	55.77	62.72	—	—	64.58	—	—	—	Anthropic	Proprietary
16	Grok 4.6	70.08	69.47	66.46	76.20	—	70.30	—	—	—	xAI	Proprietary
17	Gemini 3.1 Pro	69.95	38.92	46.29	—	79.20	66.05	—	—	—	Google	Proprietary
18	Claude Sonnet 5	69.84	65.82	64.04	—	77.50	66.59	—	—	—	Anthropic	Proprietary
19	Claude Opus 4.7 (Adaptive)	69.10	57.42	57.59	—	50.30	61.12	—	—	—	Anthropic	Proprietary
20	Claude Opus 4.6	69.00	50.07	56.57	—	59.50	61.02	—	52.60	—	Anthropic	Proprietary
21	GPT-5.2 Pro	68.79	—	—	—	—	—	—	—	—	OpenAI	Proprietary
22	Gemini 3.7 Flash	68.49	64.03	62.69	—	82.70	69.67	—	—	—	Google	Proprietary
23	Gemini 3.6 Flash	68.39	51.81	58.58	—	—	67.89	—	—	—	Google	Proprietary
24	GLM-5.3	68.37	68.42	61.47	75.80	—	61.74	—	—	—	Z.AI	Open Weight
25	Gemini 3.5 Flash	68.14	51.71	57.47	59.40	86.90	67.79	—	86.90	—	Google	Proprietary
26	GLM-5.2	68.12	58.45	60.99	—	—	60.71	—	89.80	—	Z.AI	Open Weight
27	Grok 4.5	68.10	60.54	62.15	44.80	—	69.44	—	—	—	xAI	Proprietary
28	Muse Spark	67.69	56.11	58.97	—	77.50	65.40	—	93.20	—	Meta	Proprietary
29	Grok 4.20	67.13	26.68	28.23	—	34.60	49.37	—	—	—	xAI	Proprietary
30	Qwen3.7 Max	67.08	41.24	49.32	—	—	61.74	100.00	91.10	—	Alibaba	Proprietary
31	Gemini 3 Pro	66.43	58.52	59.87	—	74.20	64.65	—	86.10	—	Google	Proprietary
32	DeepSeek V4 Pro 0813	66.38	56.51	51.86	60.30	—	59.81	—	—	—	DeepSeek	Proprietary
33	GLM-5.3-Flash	65.99	60.06	58.11	—	80.50	65.96	—	—	—	Z.AI	Open Weight
34	Kimi K2.7 Code	65.49	46.76	50.91	—	—	61.60	—	76.60	—	Moonshot AI	Open Weight
35	Ornith-1.5-397B	65.40	60.82	58.82	—	—	60.37	—	—	—	Ornith AI	Open Weight
36	Kimi K2.6	65.39	46.66	50.81	—	64.00	61.50	—	—	71.30	Moonshot AI	Open Weight
37	GPT-5.2	64.88	42.86	46.43	—	66.30	61.67	—	92.60	—	OpenAI	Proprietary
38	GPT-5.6 Luna	64.65	56.79	66.76	51.20	67.10	64.05	—	—	—	OpenAI	Proprietary
39	Qwen3.8-27B	64.52	63.42	53.85	77.40	80.00	54.05	—	84.70	—	Alibaba	Open Weight
40	dots3-note Preview	64.41	59.29	57.24	—	—	57.88	—	86.30	—	Dots Studio	Open Weight
41	GPT-5.3 Codex	64.29	61.63	62.02	—	—	65.63	—	92.60	—	OpenAI	Proprietary
42	Qwen 3.6 Max (preview)	63.70	59.74	57.58	—	—	60.22	—	—	—	Alibaba	Proprietary
43	GLM-5.1	63.24	54.26	56.39	—	—	54.77	—	93.70	64.10	Z.AI	Open Weight
44	GPT-5.1	63.03	59.37	45.00	—	—	58.07	—	89.30	—	OpenAI	Proprietary
45	Claude Sonnet 4.6	62.96	44.00	52.16	—	54.10	55.80	—	48.20	—	Anthropic	Proprietary
46	MiMo-V2.5-Pro	62.85	39.84	55.36	—	—	54.00	—	93.70	—	Xiaomi	Proprietary
47	MiMo-V2-Pro	62.75	—	55.26	—	—	53.75	—	84.00	—	Xiaomi	Proprietary
48	Gemini 3 Flash	62.20	33.99	41.54	—	—	56.92	—	66.20	—	Google	Proprietary
49	Claude Opus 4.6 (Adaptive)	61.85	52.96	56.82	—	—	—	—	—	—	Anthropic	Proprietary
50	GLM-5-Turbo	61.64	—	—	—	—	53.91	—	89.70	—	Z.AI	Proprietary
51	Qwen3.7 Plus	61.56	37.33	52.50	—	72.50	56.18	78.90	91.10	—	Alibaba	Proprietary
52	MiniMax M3	61.55	42.05	48.79	78.00	52.20	53.17	—	93.70	—	MiniMax	Open Weight
53	GLM-5	61.45	51.01	56.14	—	—	54.07	48.70	88.50	56.90	Z.AI	Open Weight
54	GPT-5.5 Pro	61.32	58.41	—	—	—	58.91	—	—	—	OpenAI	Proprietary
55	GPT-5.4 mini	61.11	39.14	42.71	—	57.20	55.62	—	89.80	—	OpenAI	Proprietary
56	Hy4 preview	61.08	58.19	59.17	—	—	56.45	—	—	—	Tencent	Open Weight
57	Hy3	60.98	47.81	58.05	—	—	56.20	—	—	—	Tencent	Open Weight
58	Qwen3.6 Plus	60.40	35.97	46.86	—	66.30	53.97	69.70	85.50	62.40	Alibaba	Proprietary
59	Inkling	60.28	38.30	41.09	—	49.20	57.38	—	85.20	—	Thinking Machines Lab	Open Weight
60	GPT-5.4 Pro	60.01	55.93	—	—	—	58.47	—	—	—	OpenAI	Proprietary
61	Grok 4.3	59.73	26.88	33.61	—	65.60	56.96	—	94.40	—	xAI	Proprietary
62	GPT-5.4 nano	59.57	34.59	37.05	—	23.80	47.44	—	93.20	—	OpenAI	Proprietary
63	Inkling-Small	59.25	36.90	43.81	—	48.80	56.52	—	89.60	—	Thinking Machines Lab	Open Weight
64	Gemini 3.5 Flash-Lite	58.99	43.03	43.38	—	76.30	52.14	—	—	—	Google	Proprietary
65	Gemini 3 Pro Deep Think	58.94	—	—	—	—	—	—	—	—	Google	Proprietary
66	Claude Opus 4.5	58.41	43.17	56.58	—	23.50	53.94	82.90	39.50	58.30	Anthropic	Proprietary
67	MiMo-V2.5	57.85	49.42	43.55	—	59.20	52.08	—	—	—	Xiaomi	Proprietary
68	MiMo-V2-Omni	57.75	—	—	—	—	49.45	—	64.10	—	Xiaomi	Proprietary
69	GLM-4.7	57.71	45.33	47.70	—	—	47.61	—	82.80	—	Z.AI	Open Weight
70	GLM-5V-Turbo	57.64	—	52.81	—	—	51.39	—	74.00	—	Z.AI	Proprietary
71	Grok 4.1	57.60	—	—	—	—	—	—	—	—	xAI	Proprietary
72	GLM-5 (Reasoning)	57.43	—	—	—	—	—	—	—	—	Z.AI	Open Weight
73	Qwen3.5 397B (Reasoning)	57.15	—	—	—	—	—	—	—	—	Alibaba	Open Weight
74	Gemini 2.5 Pro	57.12	47.71	31.73	—	—	50.55	—	57.90	—	Google	Proprietary
75	Kimi K2.5 (Reasoning)	57.00	48.18	50.66	—	—	50.66	—	—	—	Moonshot AI	Proprietary
76	DeepSeek V3.2	56.88	—	49.97	—	—	48.87	—	58.30	—	DeepSeek	Open Weight
77	Qwen3.8-Flash-Next	56.84	58.47	57.86	—	83.30	55.46	—	88.00	—	Alibaba	Open Weight
78	GPT-5.2 Instant	56.67	—	—	—	—	—	—	—	—	OpenAI	Proprietary
79	GPT-5.3 Instant	56.56	—	—	—	—	—	—	—	—	OpenAI	Proprietary
80	Qwen3.5-122B-A10B	56.37	43.62	43.57	—	57.00	48.99	36.80	93.00	—	Alibaba	Open Weight
81	GPT-5 (high)	56.30	48.83	—	—	—	—	—	—	—	OpenAI	Proprietary
82	Gemini 3.1 Flash-Lite	56.19	42.93	37.90	—	—	51.50	—	—	—	Google	Proprietary
83	Qwen3.5 Flash	56.09	—	46.66	—	—	—	—	—	—	Alibaba	Proprietary
84	DeepSeek V3.2 (Thinking)	55.80	—	47.67	—	—	—	—	—	—	DeepSeek	Open Weight
85	Qwen3 235B 2507 (Reasoning)	55.69	—	—	—	—	—	—	—	—	Alibaba	Open Weight
86	Hy3 Preview	55.34	47.71	44.33	—	—	48.82	—	54.80	—	Tencent	Open Weight
87	Qwen3.5-27B	55.34	46.73	49.98	—	—	47.22	36.80	92.80	—	Alibaba	Open Weight
88	GLM-4.5	55.26	—	—	—	—	—	—	—	—	Z.AI	Proprietary
89	MiniMax M2.7	55.19	41.19	48.69	—	—	48.68	—	93.00	—	MiniMax	Open Weight
90	MiniMax M2.5	55.09	41.09	48.59	—	—	—	—	—	—	MiniMax	Proprietary
91	Gemma 4 26B A4B	55.09	51.77	41.32	—	44.10	43.24	—	88.70	—	Google	Open Weight
92	Grok 4.1 Fast (Reasoning)	55.00	—	46.02	—	—	49.46	—	63.10	—	xAI	Proprietary
93	Qwen3.5 397B	54.70	46.78	50.19	—	63.00	50.16	69.70	—	—	Alibaba	Open Weight
94	GPT-5.3-Codex-Spark	54.60	—	—	—	—	—	—	—	—	OpenAI	Proprietary
95	GPT-5.2-Codex	54.19	50.56	51.53	—	—	56.18	—	93.70	—	OpenAI	Proprietary
96	Kimi K2.5	54.01	46.41	50.56	—	65.70	52.03	38.20	85.80	62.40	Moonshot AI	Open Weight
97	Claude Sonnet 4.5	53.82	44.22	50.60	—	—	51.16	—	—	—	Anthropic	Proprietary
98	Qwen3 235B 2507	53.73	—	—	—	—	45.67	1.00	—	—	Alibaba	Open Weight
99	Trinity-Large-Preview	53.65	—	—	—	—	—	—	—	—	Arcee AI	Open Weight
100	Qwen3.5-35B-A3B	53.39	45.02	45.47	—	—	45.40	21.10	88.80	—	Alibaba	Open Weight
101	Grok 4	53.38	—	—	—	—	52.53	—	64.40	—	xAI	Proprietary
102	Claude Opus 4.5 Thinking	52.86	—	—	—	—	54.19	—	70.00	—	Anthropic	Proprietary
103	Claude Haiku 4.5	52.79	27.00	27.20	—	—	44.23	—	—	—	Anthropic	Proprietary
104	Gemma 4 31B	52.56	45.47	42.36	—	58.40	45.77	—	92.80	—	Google	Open Weight
105	DeepSeek LLM 2.0	52.26	—	—	—	—	—	—	—	—	DeepSeek	Open Weight
106	GPT-5.1-Codex-Max	52.21	—	—	—	—	—	—	—	—	OpenAI	Proprietary
107	MAI-Thinking-1	52.00	50.30	50.39	—	—	52.28	—	94.70	—	Microsoft	Proprietary
108	GLM-4.6	51.84	—	47.06	—	—	43.99	—	42.40	—	Z.AI	Open Weight
109	Grok 3 [Beta]	51.71	—	—	—	—	—	—	—	—	xAI	Proprietary
110	Apodex 1.1	51.55	53.30	52.29	—	—	54.56	—	—	—	Apodex	Proprietary
111	Grok 4 Fast (Reasoning)	51.31	—	41.14	—	—	47.08	—	60.30	—	xAI	Proprietary
112	Gemini 2.5 Flash	50.97	—	—	—	—	45.27	—	45.30	—	Google	Proprietary
113	Nemotron 3 Nano 30B	50.71	—	—	—	—	—	—	—	—	NVIDIA	Open Weight
114	DeepSeek V3.1	50.61	—	—	—	—	44.60	—	43.80	—	DeepSeek	Open Weight
115	DeepSeek-R1	50.21	—	—	—	—	47.58	—	46.10	—	DeepSeek	Open Weight
116	Step 3.7 Flash	49.95	48.59	47.05	—	—	48.42	—	82.10	—	StepFun	Open Weight
117	Qwen3.5 Plus	49.89	47.98	—	—	—	—	—	—	—	Alibaba	Proprietary
118	Step 3.5 Flash	49.85	—	—	—	—	—	—	—	—	StepFun	Open Weight
119	GPT-5 (medium)	49.82	—	48.25	—	—	50.88	—	86.30	—	OpenAI	Proprietary
120	Qwen3.7 Flash	49.01	—	—	—	—	—	—	—	—	Alibaba	Proprietary
121	MiMo-V2-Flash	48.98	46.86	49.38	—	—	42.19	—	46.50	—	Xiaomi	Open Weight
122	DeepSeek V3.1 (Reasoning)	48.80	—	—	—	—	46.56	—	48.60	—	DeepSeek	Open Weight
123	Nemotron 3 Super 120B A12B	48.77	—	—	—	—	—	—	—	—	NVIDIA	Open Weight
124	DeepSeek Coder 2.0	48.09	—	—	—	—	—	—	—	—	DeepSeek	Open Weight
125	Seed 1.6	48.05	—	—	—	—	—	—	—	—	ByteDance	Proprietary
126	GPT-5.1-Codex	47.93	48.61	44.75	—	—	51.54	—	85.60	—	OpenAI	Proprietary
127	Nemotron 3 Super 100B	47.90	—	—	—	—	—	—	—	—	NVIDIA	Open Weight
128	o4-mini (high)	47.82	—	—	—	—	—	—	—	—	OpenAI	Proprietary
129	DeepSeekMath V2	47.71	—	—	—	—	—	—	—	—	DeepSeek	Open Weight
130	Qwen2.5-1M	47.71	—	—	—	—	—	—	—	—	Alibaba	Open Weight
131	Qwen3.6-27B	47.69	30.94	41.69	—	51.90	46.56	—	82.40	—	Alibaba	Open Weight
132	Seed-2.0-Lite	47.66	—	—	—	—	—	—	—	—	ByteDance	Proprietary
133	o3	47.55	—	—	—	—	48.95	—	87.40	—	OpenAI	Proprietary
134	Ling 3.0 Flash	47.51	41.90	40.38	—	—	44.57	—	75.60	—	InclusionAI	Open Weight
135	o3-pro	47.33	—	—	—	—	48.40	—	—	—	OpenAI	Proprietary
136	Grok 4.1 Fast	47.26	—	—	—	—	40.63	—	42.10	—	xAI	Proprietary
137	Ministral 3 14B (Reasoning)	47.18	—	—	—	—	—	—	—	—	Mistral	Open Weight
138	o3-mini	46.83	—	45.26	—	—	39.73	—	—	—	OpenAI	Proprietary
139	Grok 3 Mini	46.52	—	—	—	—	—	—	—	—	xAI	Proprietary
140	Aion-2.0	46.50	—	—	—	—	—	—	—	—	Aion Labs	Proprietary
141	GLM-4.7-Flash	46.29	—	—	—	—	—	—	—	—	Z.AI	Open Weight
142	o1	46.15	—	44.23	—	—	44.92	—	85.90	—	OpenAI	Proprietary
143	o1-preview	46.05	—	43.98	—	—	44.82	—	—	—	OpenAI	Proprietary
144	GPT-OSS 120B	45.69	40.79	43.46	—	—	42.20	—	84.30	—	OpenAI	Open Weight
145	Mistral Large 3	45.61	39.81	25.36	—	—	41.82	—	41.70	—	Mistral	Proprietary
146	GPT-5 mini	45.06	39.04	40.20	—	—	—	—	—	—	OpenAI	Proprietary
147	Muse Glimmer 30B	44.82	42.95	48.11	—	46.40	47.07	—	80.10	—	Meta	Open Weight
148	GLM-4.5-Air	43.84	—	—	—	—	39.77	—	43.50	—	Z.AI	Proprietary
149	Qwen3.6-35B-A3B	43.82	40.13	44.89	—	50.10	44.18	—	78.30	—	Alibaba	Open Weight
150	Trinity-Large-Thinking	43.40	38.03	27.31	—	—	42.02	—	67.80	—	Arcee AI	Open Weight
151	Z-1	43.05	—	—	—	—	—	—	—	—	Z	Proprietary
152	Seed 1.6 Flash	43.01	—	—	—	—	—	—	—	—	ByteDance	Proprietary
153	Mistral Small 4	42.97	37.81	40.92	—	—	40.81	—	57.30	—	Mistral	Open Weight
154	Mistral 8x7B	42.91	—	—	—	—	—	—	—	—	Mistral	Open Weight
155	Gemma 4 12B	42.72	44.88	44.04	—	26.10	38.94	—	90.10	—	Google	Open Weight
156	Moonshot v1	42.72	—	—	—	—	—	—	—	—	Moonshot AI	Proprietary
157	Command A+	42.58	40.60	43.41	—	15.10	36.56	—	90.60	—	Cohere	Open Weight
158	Seed-2.0-Mini	42.53	—	—	—	—	—	—	—	—	ByteDance	Proprietary
159	Nemotron Ultra 253B	42.34	—	—	—	—	—	—	—	—	NVIDIA	Open Weight
160	Mercury 2	42.31	38.01	24.49	—	—	—	—	—	—	Inception	Proprietary
161	Claude 4.1 Opus	42.02	42.88	42.89	—	—	—	—	—	—	Anthropic	Proprietary
162	GPT-4.1	42.02	—	41.76	—	—	39.76	—	50.50	—	OpenAI	Proprietary
163	o1-pro	41.77	—	—	—	—	—	—	—	—	OpenAI	Proprietary
164	DeepSeek V3	41.49	36.68	39.25	—	—	39.27	—	39.90	—	DeepSeek	Open Weight
165	GPT-OSS 20B	40.73	36.32	38.30	—	—	39.07	—	79.20	—	OpenAI	Open Weight
166	Qwen3 Max	40.71	—	—	—	—	41.83	—	52.00	—	Alibaba	Proprietary
167	Nemotron 3 Nano Omni 30B A3B	40.60	44.31	40.18	—	39.30	38.72	—	75.90	—	NVIDIA	Open Weight
168	Nemotron 3 Ultra	40.51	26.99	26.59	—	—	44.69	47.40	88.80	—	NVIDIA	Open Weight
169	Ling 2.6 Flash	40.30	41.80	40.28	—	—	39.68	—	46.60	—	InclusionAI	Open Weight
170	Gemma 4 E4B	39.89	43.36	39.16	—	—	37.74	—	52.10	—	Google	Open Weight
171	Sarvam 105B	39.82	—	—	—	—	40.76	—	39.40	—	Sarvam	Open Weight
172	LFM2.5-2.6B	39.60	36.79	38.77	—	—	36.66	—	47.60	—	LiquidAI	Open Weight
173	Claude 4 Sonnet	39.41	40.04	40.04	—	—	39.43	—	53.60	—	Anthropic	Proprietary
174	Gemma 4 E2B	39.27	42.86	38.59	—	—	36.38	—	44.00	—	Google	Open Weight
175	LFM2.5-8B-A1B	38.98	43.31	—	—	—	34.41	—	44.70	—	LiquidAI	Open Weight
176	GPT-4o	38.90	—	—	—	—	37.08	—	39.20	—	OpenAI	Proprietary
177	Sarvam 30B	38.62	—	—	—	—	38.71	—	29.10	—	Sarvam	Open Weight
178	Granite 4.2 8B	38.51	31.77	37.16	—	—	34.69	—	84.40	—	IBM	Open Weight
179	Ministral 3 8B (Reasoning)	38.41	—	—	—	—	—	—	—	—	Mistral	Open Weight
180	Qwen3-Omni-30B-A3B-Instruct	38.33	—	—	—	—	38.73	—	35.20	—	Alibaba	Open Weight
181	Granite-4.0-H-1B	37.90	—	—	—	—	35.58	—	28.70	—	IBM	Open Weight
182	Qwen2.5-VL-32B	37.89	—	—	—	—	—	—	—	—	Alibaba	Open Weight
183	Llama 4 Behemoth	37.84	—	—	—	—	—	—	—	—	Meta	Open Weight
184	Granite-4.0-1B	37.80	—	—	—	—	35.74	—	28.70	—	IBM	Open Weight
185	Granite-4.0-350M	37.70	—	—	—	—	34.97	—	28.70	—	IBM	Open Weight
186	Granite-4.0-H-350M	37.70	—	—	—	—	35.04	—	28.70	—	IBM	Open Weight
187	Ministral 3 3B (Reasoning)	37.56	—	—	—	—	—	—	—	—	Mistral	Open Weight
188	Qwen2.5-72B	37.37	—	—	—	—	—	—	—	—	Alibaba	Open Weight
189	Mistral 8x7B v0.2	37.19	—	—	—	—	—	—	—	—	Mistral	Open Weight
190	Ornith-1.5-35B-A3B	37.06	46.65	46.55	—	—	41.41	—	—	—	Ornith AI	Open Weight
191	GPT-5 nano	36.85	—	—	—	—	—	—	—	—	OpenAI	Proprietary
192	Llama 3.1 405B	36.55	—	—	—	—	—	—	—	—	Meta	Open Weight
193	Mistral Large 2	36.41	—	—	—	—	36.18	—	35.20	—	Mistral	Proprietary
194	Gemma 3 27B	36.26	33.04	33.32	—	—	33.58	—	36.00	—	Google	Open Weight
195	Claude 3 Opus	35.42	—	34.56	—	—	34.04	—	—	—	Anthropic	Proprietary
196	Grok Code Fast 1	34.16	—	29.87	—	—	37.20	—	48.40	—	xAI	Proprietary
197	GPT-4.1 mini	33.82	37.02	35.43	—	—	34.61	—	44.40	—	OpenAI	Proprietary
198	GPT-4o mini	33.72	33.65	31.89	—	—	31.14	—	35.00	—	OpenAI	Proprietary
199	Llama 4 Scout	33.61	31.39	31.21	—	—	33.04	—	46.00	—	Meta	Open Weight
200	Gemini 1.5 Pro	33.55	—	34.10	—	—	33.53	—	—	—	Google	Proprietary
201	Qwen2.5 Coder 32B Instruct	33.37	—	—	—	—	32.54	—	—	—	Alibaba	Open Weight
202	Phi-4	32.89	—	—	—	—	32.98	—	28.70	—	Microsoft	Open Weight
203	DeepSeek R1 Distill Qwen 32B	32.45	—	—	—	—	34.01	—	28.70	—	DeepSeek	Open Weight
204	Claude 3.5 Sonnet	31.96	—	—	—	—	—	—	—	—	Anthropic	Proprietary
205	Claude 4.1 Opus Thinking	31.61	—	—	—	—	35.95	—	66.60	—	Anthropic	Proprietary
206	Llama 3 70B	30.57	—	—	—	—	—	—	—	—	Meta	Open Weight
207	Mistral Medium 3	30.55	—	—	—	—	35.19	—	45.70	—	Mistral	Proprietary
208	Ministral 3 14B	30.27	29.31	30.17	—	—	—	—	—	—	Mistral	Open Weight
209	Mistral Medium 3.5 128B	30.13	21.86	36.86	—	—	38.98	—	84.00	—	Mistral	Open Weight
210	GPT-4.1 nano	27.15	33.37	30.89	—	—	29.62	—	36.30	—	OpenAI	Proprietary
211	Mixtral 8x22B Instruct v0.1	26.65	—	—	—	—	—	—	—	—	Mistral	Open Weight
212	GPT-4 Turbo	24.78	—	27.88	—	—	28.43	—	—	—	OpenAI	Proprietary
213	Kimi K2	23.26	—	—	—	—	33.21	—	48.60	—	Moonshot AI	Proprietary
214	Llama 4 Maverick	22.11	23.79	25.12	—	—	30.49	—	50.50	—	Meta	Open Weight
215	MiniMax M1 80k	21.51	—	—	—	—	—	—	—	—	MiniMax	Proprietary
216	Nemotron 3.5 Lightning 30B A3B NVFP4	21.01	26.06	30.22	—	—	30.19	—	72.60	—	NVIDIA	Open Weight
217	Nova Pro	19.14	—	—	—	—	26.77	—	44.20	—	Amazon	Proprietary
218	LFM2-24B-A2B	18.57	—	—	—	—	—	—	—	—	LiquidAI	Proprietary
219	Ministral 3 8B	17.87	20.97	21.06	—	—	—	—	—	—	Mistral	Open Weight
220	Claude 3 Haiku	17.55	—	—	—	—	25.09	—	41.60	—	Anthropic	Proprietary
221	LFM2.5-1.2B-Thinking	16.74	—	—	—	—	—	—	—	—	LiquidAI	Proprietary
222	LFM2.5-1.2B-Instruct	16.06	—	—	—	—	—	—	—	—	LiquidAI	Proprietary
223	Ministral 3 3B	16.04	19.80	18.93	—	—	—	—	—	—	Mistral	Open Weight
224	Nemotron-4 15B	15.90	—	—	—	—	—	—	—	—	NVIDIA	Open Weight
225	Gemini 1.0 Pro	15.66	—	—	—	—	21.14	—	—	—	Google	Proprietary
226	Mistral 7B v0.3	8.66	—	—	—	—	—	—	—	—	Mistral	Open Weight
227	Laguna M.1	3.93	23.89	29.69	—	—	18.55	—	—	—	Poolside	Proprietary
228	Laguna XS.2	1.53	22.39	24.73	—	—	17.69	—	—	—	Poolside	Open Weight

2026-09-10 AI大模型排行

猫目

热门推荐

关于

关注我们