模型评估排行榜 (Model Evaluation Leaderboard)

排名	模型	总分	代码分	截图分	视频分	分数分布 (优秀/良好/及格/差)	游戏数
🥇	gpt-5	44.99	96.58	17.64	20.74	288 676 1172	2219
🥈	o3	44.76	92.26	20.17	21.85	341 686 1127	2219
🥉	gpt-5-mini	43.48	96.69	15.71	18.02	236 655 1267	2219
4	gemini-2.5-pro	43.47	89.11	19.12	22.17	337 672 1165	2219
5	gpt-oss-120b	43.36	90.09	19.70	20.28	324 634 1209	2219
6	o4-mini	42.99	87.77	19.79	21.39	313 682 1188	2219
7	gpt-4.1-2025-04-14	42.49	91.78	17.58	18.11	263 641 1268	2219
8	Qwen3-235B-A22B-Thinking-2507	42.28	84.47	20.00	22.38	322 695 1180	2219
9	gpt-oss-20b	42.19	88.75	18.63	19.17	299 626 1263	2219
10	grok-4	42.01	83.85	19.79	22.39	327 650 1221	2219
11	gemini-2.5-flash	41.99	92.79	16.51	16.67	252 634 1304	2218
12	Qwen3-Coder-480B-A35B-Instruct	41.35	85.26	18.32	20.46	287 644 1268	2219
13	chatgpt-4o-latest	41.24	82.51	19.90	21.33	305 613 1276	2219
14	Qwen3-235B-A22B	41.23	81.27	19.79	22.64	302 668 1235	2219
15	DeepSeek-V3-0324	41.15	83.65	19.26	20.53	311 638 1248	2219
16	Qwen3-235B-A22B-Instruct-2507	41.08	85.34	18.21	19.70	308 589 1306	2219
17	DeepSeek-V3.1	40.85	83.08	19.27	20.20	296 611 1287	2219
18	claude-sonnet-4-20250514-thinking	40.51	90.27	14.38	16.88	204 624 1355	2219
19	Qwen3-32B	40.38	81.58	18.92	20.64	274 642 1295	2219
20	Seed-OSS-36B-Instruct	40.31	88.28	16.42	16.23	234 585 1375	2219
21	claude-sonnet-4-20250514	40.24	87.68	15.66	17.38	207 616 1360	2219
22	DeepSeek-R1	40.10	81.00	19.22	20.07	278 607 1319	2219
23	GLM-4.5	40.01	84.72	16.98	18.32	216 661 1311	2219
24	Qwen3-30B-A3B-Thinking-2507	39.99	80.65	18.93	20.39	279 589 1338	2219
25	QwQ-32B	39.60	79.69	18.55	20.57	268 610 1331	2219
26	Qwen3-30B-A3B	39.57	78.36	19.66	20.68	274 597 1339	2219
27	GLM-4.5-Air	39.41	85.41	16.31	16.52	230 533 1433	2219
28	Qwen3-Coder-30B-A3B-Instruct	39.02	83.78	16.59	16.70	226 543 1428	2219
29	Qwen3-14B	38.78	79.09	18.41	18.84	245 567 1398	2219
30	DeepSeek-R1-0528	38.71	88.07	13.44	14.61	174 544 1469	2219
31	Qwen3-30B-A3B-Instruct-2507	38.57	81.36	16.54	17.83	223 571 1414	2219
32	o3-mini-2025-01-31	38.18	89.29	11.91	13.34	204 417 1572	2219
33	gpt-4o-2024-11-20	37.59	76.62	17.51	18.64	224 531 1452	2219
34	Qwen3-8B	36.91	76.17	17.24	17.31	187 546 1480	2218
35	DeepSeek-V3	36.65	73.37	17.65	18.94	204 564 1447	2218
36	DeepSeek-R1-Distill-Llama-70B	35.28	74.06	15.76	16.04	188 448 1579	2219
37	Zhihu-ai-Zhi-Create-Qwen3-32B	35.15	75.83	15.19	14.42	184 426 1606	2219
38	Qwen2.5-72B-Instruct	34.58	73.22	14.66	15.87	174 449 1593	2219
39	Qwen2.5-Coder-32B-Instruct	34.42	74.55	13.77	14.94	167 425 1618	2219
40	Qwen3-4B	34.41	72.66	15.11	15.46	144 464 1610	2219
41	Qwen3-4B-Thinking-2507	34.28	69.98	16.13	16.75	168 465 1584	2219
42	gpt-4o-mini-2024-07-18	33.90	70.36	15.52	15.81	134 459 1622	2219
43	Seed-Coder-8B-Instruct	33.87	73.22	13.98	14.41	137 422 1656	2219
44	DeepSeek-R1-Distill-Qwen-32B	33.40	71.87	14.39	13.95	145 411 1663	2219
45	Qwen2.5-32B-Instruct	31.84	66.38	14.04	15.11	127 389 1701	2219
46	Qwen2.5-14B-Instruct	30.27	66.42	11.43	12.97	348 1779	2219
47	Qwen2.5-Coder-14B-Instruct	30.18	68.50	10.89	11.17	327 1804	2218
48	Llama-3.1-8B-Instruct	29.39	62.86	12.96	12.34	319 1815	2219
49	Qwen2.5-Coder-7B-Instruct	27.58	63.91	9.09	9.73	250 1899	2218
50	DeepSeek-R1-Distill-Qwen-14B	27.43	65.33	8.68	8.28	198 1943	2219
51	Qwen2.5-7B-Instruct	26.09	59.81	9.22	9.24	230 1937	2219
52	Qwen3-1.7B	25.03	57.32	9.09	8.69	216 1959	2218
53	Llama-3.2-3B-Instruct	24.50	55.51	9.49	8.50	210 1963	2219
54	deepseek-coder-7b-instruct-v1.5	24.00	53.78	9.02	9.21	229 1952	2219
55	kimi-k2-0905-preview	23.47	66.26	1.95	2.21	2135	2215
56	Hunyuan-7B-Instruct	20.96	57.80	2.84	2.25	2148	2219
57	Qwen2.5-Coder-3B-Instruct	20.36	53.81	4.37	2.89	2124	2219
58	OpenCoder-8B-Instruct	20.11	49.02	6.55	4.75	152 2060	2219
59	Qwen2.5-3B-Instruct	18.70	46.92	5.04	4.14	2116	2219
60	Qwen2.5-Coder-1.5B-Instruct	17.26	46.59	3.95	1.24	2159	2219
61	OpenCoder-1.5B-Instruct	16.86	43.09	5.61	1.87	2108	2219
62	Llama-3.2-1B-Instruct	15.51	40.28	4.33	1.92	2158	2219
63	Qwen2.5-1.5B-Instruct	15.19	40.07	3.73	1.78	2154	2219
64	DeepSeek-R1-Distill-Llama-8B	15.08	42.45	1.74	1.04	2187	2219
65	DeepSeek-R1-0528-Qwen3-8B	14.02	41.29	0.43	0.33	2207	2218
66	Qwen3-0.6B	13.67	35.11	4.82	1.07	2160	2219
67	Qwen2.5-Coder-0.5B-Instruct	12.75	34.64	3.57	0.04	2179	2218
68	DeepSeek-R1-Distill-Qwen-7B	12.07	35.78	0.30	0.12	2216	2219
69	Qwen2.5-0.5B-Instruct	10.86	30.81	1.73	0.04	2200	2216
70	DeepSeek-R1-Distill-Qwen-1.5B	8.48	25.40	0.04	0.00	2218	2219

🏆 模型评估排行榜

📁 Resources & Links