排行榜

每个来源独立计量；下方展示本次抓取的全部记录，可在站内搜索、筛选与排序。快照不会自动随原站变化。

ARENA / TEXT · STYLE CONTROL

AI 模型文本榜

Arena 文本对战综合榜：完整展示 latest 快照中 overall 类别的 402 条模型记录，包括专有与开放权重模型；评分不等同于能力的绝对值。

来源与方法 ↗

榜单发布 2026-09-13抓取 2026年9月24日 05:19 UTC数据：Arena，CC BY 4.0 ↗；本站按 overall 类别筛选，评分显示两位小数；原榜未标注机构的模型仅补充可核对的发布方标志

搜索模型 / 机构机构排序

402 / 402 条记录

表格可横向滚动；原榜置信区间和模型许可一并保留。

Arena 文本综合榜全部模型
原榜	模型	机构	Arena 评分	95% 区间	对战票数	模型许可	方差
#1	claude-fable-5	anthropic	1505.68	1500.92–1510.45	30,057	Proprietary	5.91
#2	claude-opus-4-6-high	anthropic	1504.56	1501.04–1508.08	71,993	Proprietary	3.22
#3	claude-opus-4-7-high	anthropic	1501.75	1497.87–1505.63	60,002	Proprietary	3.92
#4	muse-spark-1.2 (xHigh)	meta	1499.59	1489.08–1510.10	3,227	Proprietary	28.74
#5	claude-fable-5.1-max	anthropic	1498.47	1490.31–1506.64	5,783	Proprietary	17.36
#6	claude-opus-4-6	anthropic	1497.35	1493.93–1500.77	75,878	Proprietary	3.05
#7	claude-opus-4-7	anthropic	1494.45	1490.58–1498.32	61,128	Proprietary	3.90
#8	muse-spark-1.3-max	meta	1493.11	1484.27–1501.96	4,723	Proprietary	20.36
#9	gemini-3.8-flash-high	google	1493.01	1484.42–1501.59	5,076	Proprietary	19.18
#10	claude-opus-5-high	anthropic	1492.90	1488.67–1497.13	42,617	Proprietary	4.66
#11	muse-spark-1.1	meta	1492.56	1487.70–1497.41	27,615	Proprietary	6.14
#12	gemini-3.7-flash-high	google	1489.82	1481.62–1498.02	5,640	Proprietary	17.50
#13	muse-spark	meta	1488.05	1482.21–1493.90	13,565	Proprietary	8.89
#14	claude-opus-5-max	anthropic	1487.36	1481.87–1492.86	20,706	Proprietary	7.85
#15	gemini-3.1-pro-preview	google	1486.81	1483.65–1489.97	106,951	Proprietary	2.60
#16	gemini-3-pro	google	1485.48	1481.63–1489.34	40,654	Proprietary	3.87
#17	kimi-k3-max	moonshot	1484.77	1479.54–1489.99	20,987	Kimi K3 license	7.11
#18	gpt-5.6-sol-xhigh	openai	1483.47	1478.58–1488.37	27,069	Proprietary	6.23
#19	glm-5.3-max	zai	1483.02	1476.58–1489.46	10,960	MIT	10.80
#20	gpt-5.5-high	openai	1481.76	1477.94–1485.58	64,924	Proprietary	3.80
#21	claude-opus-4-8-high	anthropic	1481.38	1477.22–1485.53	52,535	Proprietary	4.49
#22	qwen3.8-max	alibaba	1480.56	1474.77–1486.35	16,670	Proprietary	8.73
#23	gemini-3.6-flash-high	google	1479.89	1475.04–1484.74	26,445	Proprietary	6.13
#24	gpt-6-astra-max	openai	1479.77	1468.12–1491.42	2,693	Proprietary	35.33
#25	gemini-3.5-flash-high	google	1477.68	1473.49–1481.87	38,257	Proprietary	4.58
#26	gpt-5.4-high	openai	1476.39	1472.51–1480.26	60,537	Proprietary	3.91
#27	gpt-5.2-chat-latest-20260210	openai	1476.26	1472.12–1480.39	34,176	Proprietary	4.45
#28	gpt-5.5	openai	1476.08	1472.29–1479.88	66,317	Proprietary	3.75
#29	glm-5.3-flash	zai	1475.40	1468.84–1481.97	10,038	MIT	11.21
#30	grok-4.20-beta1	xai	1474.50	1469.85–1479.16	26,599	Proprietary	5.63
#31	gemini-3.5-flash-medium	google	1473.94	1469.44–1478.44	36,627	Proprietary	5.27
#32	gpt-5.5-instant	openai	1473.82	1468.71–1478.93	25,850	Proprietary	6.80
#33	gemini-3-flash	google	1473.69	1469.27–1478.11	30,225	Proprietary	5.08
#34	qwen3.7-max-preview	alibaba	1473.23	1463.22–1483.23	3,705	Proprietary	26.05
#35	claude-opus-4-8	anthropic	1473.22	1469.08–1477.36	53,446	Proprietary	4.46
#36	claude-opus-4-5-20251101-high-32k	anthropic	1472.79	1468.89–1476.70	36,239	Proprietary	3.97
#37	claude-sonnet-4-6	anthropic	1472.57	1468.97–1476.18	66,208	Proprietary	3.39
#38	glm-5.2-max	zai	1472.10	1467.48–1476.71	36,798	MIT	5.55
#39	grok-4.20-beta-0309-reasoning	xai	1471.48	1467.70–1475.25	62,168	Proprietary	3.71
#40	grok-4.20-multi-agent-beta-0309	xai	1470.15	1466.36–1473.95	60,777	Proprietary	3.75
#41	claude-opus-4-5-20251101	anthropic	1469.47	1466.27–1472.66	70,013	Proprietary	2.66
#42	grok-4.5	xai	1468.13	1463.44–1472.83	30,103	Proprietary	5.74
#43	ernie-5.1	baidu	1467.89	1463.23–1472.54	37,058	Proprietary	5.64
#44	mimo-v2.5-pro	xiaomi	1467.42	1463.59–1471.25	60,919	MIT	3.82
#45	gpt-5.6-terra-xhigh	openai	1466.04	1461.21–1470.87	28,119	Proprietary	6.07
#46	gpt-5.4	openai	1465.66	1461.85–1469.46	63,526	Proprietary	3.76
#47	glm-5.1	zai	1465.51	1461.64–1469.38	48,901	MIT	3.90
#48	grok-4.1-thinking	xai	1465.45	1462.20–1468.69	64,152	Proprietary	2.74
#49	qwen3.5-max-preview	alibaba	1465.23	1460.26–1470.20	21,476	Proprietary	6.43
#50	deepseek-v4-pro-high-20260813	deepseek	1463.36	1456.52–1470.20	9,008	MIT	12.17
#51	claude-sonnet-5-high	anthropic	1461.21	1456.62–1465.79	35,301	Proprietary	5.47
#52	kimi-k2.6	moonshot	1460.44	1455.91–1464.98	37,502	Modified MIT	5.35
#53	qwen3.6-max-preview	alibaba	1460.01	1451.65–1468.37	5,186	Proprietary	18.19
#54	grok-4.1	xai	1459.18	1455.96–1462.40	66,340	Proprietary	2.70
#55	gemini-3-flash (thinking-minimal)	google	1458.32	1455.22–1461.42	85,799	Proprietary	2.51
#56	glm-5	zai	1457.64	1453.30–1461.99	27,605	MIT	4.92
#57	deepseek-v4-pro	deepseek	1457.34	1453.34–1461.34	54,130	MIT	4.17
#58	claude-sonnet-4-5-20250929-high-32k	anthropic	1456.20	1453.40–1459.01	80,980	Proprietary	2.05
#59	gemini-3.5-flash-lite	google	1455.88	1450.91–1460.84	26,165	Proprietary	6.41
#60	hy3	tencent	1455.78	1448.57–1462.99	8,047	Apache 2.0	13.54
#61	dola-seed-2.0-pro	bytedance	1455.57	1452.07–1459.06	74,159	Proprietary	3.17
#62	qwen3.7-plus	alibaba	1455.57	1451.11–1460.02	39,334	Proprietary	5.17
#63	grok-4.6-high	xai	1455.56	1449.75–1461.36	15,521	Proprietary	8.78
#64	gpt-5.1-high	openai	1455.13	1451.35–1458.90	40,284	Proprietary	3.71
#65	claude-sonnet-4-5-20250929	anthropic	1455.09	1452.17–1458.00	79,628	Proprietary	2.20
#66	deepseek-v4-pro-high-preview	deepseek	1454.71	1450.63–1458.79	51,485	MIT	4.32
#67	gpt-5.6-luna-xhigh	openai	1452.45	1447.64–1457.26	28,547	Proprietary	6.02
#68	gemma-4-31b	google	1451.09	1443.50–1458.67	5,894	Apache 2.0	14.96
#69	kimi-k2.5-thinking	moonshot	1450.35	1446.99–1453.71	70,513	Modified MIT	2.94
#70	claude-opus-4-1-20250805-thinking-16k	anthropic	1449.85	1446.38–1453.33	48,811	Proprietary	3.14
#71	gpt-5.3-chat-latest	openai	1449.17	1444.84–1453.51	32,727	Proprietary	4.89
#72	ernie-5.0-preview-1203	baidu	1448.86	1442.29–1455.43	9,561	Proprietary	11.24
#73	mimo-v2-pro	xiaomi	1448.40	1443.66–1453.15	24,365	Proprietary	5.87
#74	gpt-5.4-mini-high	openai	1448.11	1444.27–1451.96	59,387	Proprietary	3.85
#75	claude-opus-4-1-20250805	anthropic	1447.72	1444.72–1450.72	75,913	Proprietary	2.34
#76	ernie-5.0-0110	baidu	1446.11	1442.13–1450.09	34,860	Proprietary	4.12
#77	gemini-2.5-pro	google	1445.50	1443.01–1447.99	122,554	Proprietary	1.62
#78	gpt-4.5-preview-2025-02-27	openai	1444.65	1438.94–1450.36	14,547	Proprietary	8.49
#79	qwen3.6-plus	alibaba	1443.27	1439.04–1447.51	45,319	Proprietary	4.67
#80	chatgpt-4o-latest-20250326	openai	1443.10	1440.27–1445.93	80,677	Proprietary	2.09
#81	grok-4.3	xai	1442.65	1438.92–1446.38	66,801	Proprietary	3.63
#82	qwen3.5-397b-a17b	alibaba	1441.83	1438.52–1445.14	77,007	Apache 2.0	2.85
#83	glm-4.7	zai	1441.73	1435.61–1447.86	11,893	MIT	9.77
#84	minimax-m3	minimax	1441.25	1437.04–1445.46	48,540	MiniMax Community License	4.62
#85	inkling	thinky	1440.12	1435.19–1445.05	25,922	Apache 2.0	6.32
#86	gpt-5.1	openai	1439.00	1435.35–1442.65	42,982	Proprietary	3.47
#87	deepseek-v4-flash-high-preview	deepseek	1437.94	1433.73–1442.14	48,597	MIT	4.61
#88	gemma-4-26b-a4b	google	1437.91	1430.29–1445.53	5,804	Apache 2.0	15.12
#89	gpt-5.2-high	openai	1437.58	1433.96–1441.21	47,538	Proprietary	3.42
#90	qwen3.8-27b	alibaba	1437.10	1430.63–1443.56	10,697	Apache 2.0	10.87
#91	gpt-5.2	openai	1435.84	1432.60–1439.09	78,967	Proprietary	2.74
#92	deepseek-v4-flash	deepseek	1435.57	1431.42–1439.71	48,887	MIT	4.47
#93	longcat-flash-chat-2602-exp	meituan	1435.49	1430.86–1440.11	27,962	Proprietary	5.57
#94	qwen3-max-preview	alibaba	1434.68	1430.15–1439.22	27,194	Proprietary	5.35
#95	gpt-5-high	openai	1434.65	1430.13–1439.18	31,428	Proprietary	5.34
#96	mimo-v2.5	xiaomi	1433.61	1429.28–1437.95	44,466	MIT	4.90
#97	glm-5v-turbo	zai	1432.68	1425.96–1439.39	9,361	Proprietary	11.74
#98	gemini-3.1-flash-lite-preview	google	1432.36	1428.64–1436.08	60,405	Proprietary	3.61
#99	o3-2025-04-16	openai	1431.55	1427.91–1435.20	58,579	Proprietary	3.46
#100	kimi-k2.5-instant	moonshot	1430.91	1424.27–1437.55	7,987	Modified MIT	11.49
#101	grok-4-1-fast-reasoning	xai	1430.26	1427.00–1433.52	55,363	Proprietary	2.76
#102	kimi-k2-thinking-turbo	moonshot	1430.13	1426.99–1433.28	61,088	Modified MIT	2.58
#103	mimo-v2-omni	xiaomi	1429.76	1424.07–1435.45	19,414	Proprietary	8.44
#104	muse-glimmer	meta	1427.16	1417.18–1437.13	3,666	Apache-2.0	25.91
#105	gpt-5-chat	openai	1426.81	1422.49–1431.13	30,957	Proprietary	4.86
#106	mistral-medium-3.5	mistral	1426.35	1419.79–1432.90	10,996	Modified MIT	11.20
#107	claude-opus-4-20250514-thinking-16k	anthropic	1425.99	1421.58–1430.40	35,735	Proprietary	5.06
#108	amazon-nova-experimental-chat-26-02-10	amazon	1425.92	1416.04–1435.79	3,369	Proprietary	25.37
#109	nvidia-nemotron-3-ultra-550b-a55b-nvfp4	nvidia	1425.59	1418.58–1432.60	10,693	OpenMDW-1.1	12.78
#110	deepseek-v3.2-exp-thinking	deepseek	1425.28	1418.69–1431.87	8,905	MIT	11.31
#111	deepseek-v3.2	deepseek	1425.18	1421.63–1428.73	46,458	MIT	3.28
#112	glm-4.6	zai	1424.74	1420.81–1428.68	35,061	MIT	4.03
#113	qwen3-max-2025-09-23	alibaba	1423.74	1417.22–1430.25	8,939	Proprietary	11.05
#114	qwen3-235b-a22b-instruct-2507	alibaba	1423.03	1420.44–1425.62	95,670	Apache 2.0	1.74
#115	deepseek-v3.2-thinking	deepseek	1422.43	1418.78–1426.07	40,499	MIT	3.45
#116	deepseek-v3.2-exp	deepseek	1422.31	1415.88–1428.73	11,760	MIT	10.76
#117	deepseek-r1-0528	deepseek	1421.37	1415.68–1427.06	18,091	MIT	8.43
#118	ernie-5.0-preview-1022	baidu	1418.81	1409.98–1427.64	4,663	Proprietary	20.28
#119	grok-4-fast-chat	xai	1418.36	1410.55–1426.16	6,335	Proprietary	15.87
#120	kimi-k2-0905-preview	moonshot	1418.06	1411.52–1424.59	11,568	Modified MIT	11.11
#121	kimi-k2-0711-preview	moonshot	1417.77	1412.86–1422.68	27,029	Modified MIT	6.27
#122	deepseek-v3.1-terminus-thinking	deepseek	1417.43	1407.32–1427.55	3,369	MIT	26.62
#123	deepseek-v3.1	deepseek	1417.31	1411.25–1423.37	14,607	MIT	9.56
#124	qwen3.5-122b-a10b	alibaba	1416.78	1412.40–1421.15	28,359	Apache 2.0	4.99
#125	deepseek-v3.1-thinking	deepseek	1415.98	1409.31–1422.66	11,460	MIT	11.60
#126	minimax-m2.7	minimax	1415.17	1411.57–1418.78	69,169	Modified MIT	3.38
#127	deepseek-v3.1-terminus	deepseek	1414.75	1405.03–1424.46	3,611	MIT	24.56
#128	amazon-nova-experimental-chat-26-01-10	amazon	1414.58	1404.63–1424.53	3,346	Proprietary	25.78
#129	claude-haiku-4-5-20251001	anthropic	1414.55	1412.03–1417.07	129,278	Proprietary	1.65
#130	gpt-4.1-2025-04-14	openai	1414.47	1410.75–1418.20	49,941	Proprietary	3.61
#131	qwen3-vl-235b-a22b-instruct	alibaba	1414.07	1407.50–1420.63	11,219	Apache 2.0	11.22
#132	claude-opus-4-20250514	anthropic	1413.88	1409.54–1418.21	42,845	Proprietary	4.89
#133	mistral-large-3	mistral	1412.99	1409.97–1416.02	69,028	Apache 2.0	2.38
#134	hunyuan-hy3-preview	tencent	1412.69	1405.15–1420.23	6,614	tencent-hunyuan-community	14.80
#135	grok-3-preview-02-24	xai	1411.53	1407.18–1415.88	32,414	Proprietary	4.92
#136	glm-4.5	zai	1411.30	1406.36–1416.24	23,712	MIT	6.35
#137	grok-4-0709	xai	1411.05	1407.09–1415.01	39,297	Proprietary	4.08
#138	gemini-2.5-flash	google	1409.69	1407.22–1412.16	122,732	Proprietary	1.59
#139	mistral-medium-2508	mistral	1408.62	1405.98–1411.26	92,697	Proprietary	1.81
#140	qwen3.5-27b	alibaba	1407.88	1403.48–1412.28	27,227	Apache 2.0	5.04
#141	Inkling Small	thinky	1404.61	1399.15–1410.07	18,844	Apache 2.0	7.77
#142	grok-4-fast-reasoning	xai	1404.56	1399.47–1409.65	17,975	Proprietary	6.74
#143	gemini-2.5-flash-preview-09-2025	google	1403.57	1399.49–1407.65	32,364	Proprietary	4.34
#144	qwen3-235b-a22b-no-thinking	alibaba	1402.52	1397.99–1407.06	37,389	Apache 2.0	5.36
#145	o1-2024-12-17	openai	1402.13	1397.66–1406.60	27,807	Proprietary	5.21
#146	gpt-5.4-nano-high	openai	1402.08	1398.20–1405.95	58,424	Proprietary	3.91
#147	longcat-flash-chat	meituan	1401.34	1394.92–1407.76	11,154	MIT	10.74
#148	claude-sonnet-4-20250514-thinking-32k	anthropic	1400.92	1396.46–1405.38	33,944	Proprietary	5.18
#149	qwen3-235b-a22b-thinking-2507	alibaba	1400.06	1393.47–1406.64	8,796	Apache 2.0	11.29
#150	qwen3-next-80b-a3b-instruct	alibaba	1399.33	1394.53–1404.13	22,529	Apache 2.0	5.99
#151	deepseek-r1	deepseek	1398.04	1393.14–1402.93	18,524	MIT	6.23
#152	qwen3.5-flash	alibaba	1396.72	1392.96–1400.48	58,193	Proprietary	3.68
#153	deepseek-v3-0324	deepseek	1395.73	1391.83–1399.63	44,787	MIT	3.96
#154	qwen3-vl-235b-a22b-thinking	alibaba	1395.15	1388.29–1402.01	7,780	Apache 2.0	12.23
#155	hunyuan-vision-1.5-thinking	tencent	1395.14	1382.75–1407.54	2,171	Proprietary	39.98
#156	qwen3.5-35b-a3b	alibaba	1394.97	1390.64–1399.29	29,043	Apache 2.0	4.86
#157	amazon-nova-experimental-chat-12-10	amazon	1394.58	1384.99–1404.16	3,632	Proprietary	23.92
#158	step-3.5-flash	stepfun	1394.13	1390.50–1397.77	57,137	Apache 2.0	3.45
#159	mimo-v2-flash (non-thinking)	xiaomi	1391.87	1388.31–1395.42	46,116	MIT	3.29
#160	o4-mini-2025-04-16	openai	1390.81	1386.81–1394.82	44,639	Proprietary	4.18
#161	claude-sonnet-4-20250514	anthropic	1390.69	1386.28–1395.10	38,966	Proprietary	5.06
#162	minimax-m2.5	minimax	1390.32	1386.34–1394.30	40,843	Modified MIT	4.12
#163	gpt-5-mini-high	openai	1389.56	1384.89–1394.23	26,569	Proprietary	5.67
#164	o1-preview	openai	1388.37	1383.36–1393.38	31,122	Proprietary	6.54
#165	claude-3-7-sonnet-20250219-thinking-32k	anthropic	1388.11	1383.90–1392.32	38,081	Proprietary	4.61
#166	qwen3-coder-480b-a35b-instruct	alibaba	1387.51	1382.52–1392.49	25,089	Apache 2.0	6.48
#167	hunyuan-t1-20250711	tencent	1387.47	1378.75–1396.19	4,580	Proprietary	19.80
#168	mistral-medium-2505	mistral	1387.40	1382.67–1392.13	32,662	Proprietary	5.82
#169	mimo-v2-flash (thinking)	xiaomi	1386.14	1379.86–1392.42	10,739	MIT	10.27
#170	minimax-m2.1-preview	minimax	1383.79	1378.53–1389.05	16,622	MIT	7.21
#171	hunyuan-turbos-20250416	tencent	1382.56	1376.06–1389.07	10,586	Proprietary	11.02
#172	gpt-4.1-mini-2025-04-14	openai	1382.52	1378.18–1386.85	38,631	Proprietary	4.89
#173	qwen3-30b-a3b-instruct-2507	alibaba	1382.34	1377.42–1387.26	23,222	Apache 2.0	6.30
#174	gemini-2.5-flash-lite-preview-09-2025-no-thinking	google	1379.59	1376.14–1383.05	46,382	Proprietary	3.11
#175	glm-4.6v	zai	1378.46	1367.13–1389.79	2,762	MIT	33.41
#176	trinity-large-preview	未标注	1378.44	1374.09–1382.80	29,742	Apache 2.0	4.94
#177	solar-pro4	upstage	1376.41	1364.17–1388.64	2,420	Proprietary	38.98
#178	qwen3-235b-a22b	alibaba	1374.89	1370.18–1379.61	25,865	Apache 2.0	5.78
#179	gemini-2.5-flash-lite-preview-06-17-thinking	google	1374.63	1370.09–1379.17	32,216	Proprietary	5.36
#180	claude-3-5-sonnet-20241022	anthropic	1374.01	1370.80–1377.22	87,694	Proprietary	2.68
#181	qwen2.5-max	alibaba	1373.89	1369.77–1378.01	32,417	Proprietary	4.42
#182	glm-4.5-air	zai	1373.21	1368.93–1377.48	30,367	MIT	4.76
#183	claude-3-7-sonnet-20250219	anthropic	1372.20	1368.22–1376.17	42,448	Proprietary	4.11
#184	qwen3-next-80b-a3b-thinking	alibaba	1368.96	1363.08–1374.84	13,380	Apache 2.0	9.01
#185	trinity-large-thinking	未标注	1368.96	1364.29–1373.64	28,967	Apache 2.0	5.69
#186	glm-4.7-flash	zai	1366.26	1360.44–1372.09	11,491	MIT	8.83
#187	gemma-3-27b-it	google	1365.19	1361.47–1368.90	46,422	Gemma	3.59
#188	amazon-nova-experimental-chat-11-10	amazon	1365.08	1360.73–1369.44	25,086	Proprietary	4.93
#189	minimax-m1	minimax	1363.92	1359.60–1368.23	34,030	Apache 2.0	4.84
#190	grok-3-mini-high	xai	1363.52	1358.03–1369.00	15,536	Proprietary	7.83
#191	o3-mini-high	openai	1363.46	1358.19–1368.73	18,589	Proprietary	7.22
#192	nvidia-nemotron-3-super-120b-a12b	nvidia	1360.54	1353.26–1367.82	7,458	NVIDIA Open Model	13.80
#193	gemini-2.0-flash-001	google	1359.87	1356.06–1363.67	43,351	Proprietary	3.77
#194	deepseek-v3	deepseek	1358.36	1353.61–1363.12	21,770	DeepSeek	5.89
#195	grok-3-mini-beta	xai	1357.95	1352.81–1363.09	20,837	Proprietary	6.89
#196	mistral-small-2506	mistral	1356.72	1351.49–1361.95	17,391	Apache 2.0	7.12
#197	intellect-3	未标注	1356.25	1347.90–1364.60	5,279	MIT	18.16
#198	command-a-03-2025	cohere	1353.62	1350.14–1357.10	55,449	CC-BY-NC-4.0	3.15
#199	gemini-2.0-flash-lite-preview-02-05	google	1353.39	1349.07–1357.72	24,955	Proprietary	4.86
#200	glm-4.5v	zai	1352.55	1344.11–1361.00	4,817	MIT	18.55
#201	gpt-oss-120b	openai	1352.28	1347.89–1356.67	29,955	Apache 2.0	5.02
#202	nvidia-nemotron-3.5-lightning-30b-a3b-nvfp4	nvidia	1351.48	1344.15–1358.80	8,338	OpenMDW-1.1	13.96
#203	gemini-1.5-pro-002	google	1351.06	1347.66–1354.47	55,606	Proprietary	3.01
#204	amazon-nova-experimental-chat-10-20	amazon	1349.52	1343.32–1355.72	11,368	Proprietary	10.01
#205	step-3	stepfun	1349.39	1341.86–1356.91	6,273	Apache 2.0	14.73
#206	hunyuan-turbos-20250226	tencent	1348.78	1337.07–1360.49	2,220	Proprietary	35.72
#207	o3-mini	openai	1347.74	1344.18–1351.29	56,655	Proprietary	3.29
#208	llama-3.1-nemotron-ultra-253b-v1	nvidia	1347.52	1335.84–1359.20	2,549	Nvidia Open Model	35.53
#209	qwen3-32b	alibaba	1346.92	1337.46–1356.38	3,926	Apache 2.0	23.29
#210	amazon-nova-experimental-chat-10-09	amazon	1346.84	1335.89–1357.78	2,775	Proprietary	31.18
#211	mercury-2	inception-ai	1346.27	1335.55–1356.99	3,042	Proprietary	29.91
#212	gpt-4o-2024-05-13	openai	1345.94	1342.47–1349.41	112,881	Proprietary	3.13
#213	qwen-plus-0125	alibaba	1345.91	1337.57–1354.24	5,819	Proprietary	18.09
#214	minimax-m2	minimax	1345.87	1338.11–1353.63	6,753	Apache 2.0	15.68
#215	ling-flash-2.0	ant-group	1343.99	1336.66–1351.32	6,850	MIT	14.00
#216	nvidia-llama-3.3-nemotron-super-49b-v1.5	nvidia	1343.14	1333.12–1353.16	3,313	Nvidia Open	26.14
#217	claude-3-5-sonnet-20240620	anthropic	1342.99	1339.54–1346.44	82,419	Proprietary	3.10
#218	glm-4-plus-0111	zai	1342.58	1334.16–1351.00	5,760	Proprietary	18.45
#219	granite-4.2-30b	ibm	1341.88	1331.20–1352.57	3,255	Apache 2.0	29.71
#220	gemma-3-12b-it	google	1341.68	1332.17–1351.19	3,829	Gemma	23.56
#221	hunyuan-turbo-0110	tencent	1340.60	1329.08–1352.13	2,290	Proprietary	34.57
#222	gpt-5-nano-high	openai	1337.47	1330.54–1344.39	8,128	Proprietary	12.47
#223	o1-mini	openai	1336.79	1333.12–1340.45	51,981	Proprietary	3.49
#224	nova-2-lite	amazon	1335.91	1329.77–1342.06	12,120	Proprietary	9.83
#225	qwq-32b	alibaba	1335.80	1331.33–1340.26	25,090	Apache 2.0	5.19
#226	gemini-advanced-0514	google	1335.65	1330.49–1340.81	50,148	Proprietary	6.93
#227	grok-2-2024-08-13	xai	1335.53	1331.84–1339.22	63,498	Proprietary	3.55
#228	gpt-4o-2024-08-06	openai	1335.35	1331.14–1339.57	45,499	Proprietary	4.62
#229	llama-3.1-405b-instruct-bf16	meta	1335.16	1331.43–1338.89	41,375	Llama 3.1 Community	3.62
#230	step-2-16k-exp-202412	stepfun	1333.54	1324.95–1342.13	4,833	Proprietary	19.21
#231	llama-3.1-405b-instruct-fp8	meta	1333.29	1329.69–1336.90	59,656	Llama 3.1 Community	3.39
#232	olmo-3.1-32b-instruct	allenai	1329.96	1323.74–1336.17	11,452	Apache 2.0	10.06
#233	yi-lightning	未标注	1328.21	1323.26–1333.16	27,332	Proprietary	6.37
#234	molmo-2-8b	allenai	1328.00	1306.62–1349.39	791	Apache 2.0	119.04
#235	llama-3.3-nemotron-49b-super-v1	nvidia	1327.74	1315.60–1339.88	2,218	Nvidia	38.35
#236	llama-4-maverick-17b-128e-instruct	meta	1326.87	1322.59–1331.15	39,359	Llama 4	4.77
#237	qwen3-30b-a3b	alibaba	1326.87	1322.11–1331.63	26,089	Apache 2.0	5.89
#238	hunyuan-large-2025-02-10	tencent	1326.23	1316.49–1335.96	3,738	Proprietary	24.66
#239	claude-3-5-haiku-20241022	anthropic	1324.31	1321.05–1327.58	68,964	Proprietary	2.77
#240	gpt-4-turbo-2024-04-09	openai	1324.26	1320.33–1328.19	98,114	Proprietary	4.02
#241	gemini-1.5-pro-001	google	1323.81	1319.82–1327.80	79,138	Proprietary	4.14
#242	deepseek-v2.5-1210	deepseek	1323.32	1315.05–1331.59	6,795	DeepSeek	17.82
#243	gpt-4.1-nano-2025-04-14	openai	1322.21	1314.48–1329.94	6,103	Proprietary	15.56
#244	claude-3-opus-20240229	anthropic	1321.92	1318.85–1325.00	194,909	Proprietary	2.47
#245	llama-4-scout-17b-16e-instruct	meta	1321.46	1316.72–1326.20	29,740	Llama	5.85
#246	ring-flash-2.0	ant-group	1320.93	1313.61–1328.24	6,899	MIT	13.94
#247	step-1o-turbo-202506	stepfun	1320.11	1313.21–1327.00	8,858	Proprietary	12.36
#248	glm-4-plus	zai	1319.30	1314.37–1324.24	26,126	Proprietary	6.34
#249	qwen-max-0919	alibaba	1318.00	1312.30–1323.71	16,478	Qwen	8.47
#250	llama-3.3-70b-instruct	meta	1317.60	1314.10–1321.11	54,412	Llama-3.3	3.20
#251	gpt-4o-mini-2024-07-18	openai	1317.59	1314.04–1321.14	68,697	Proprietary	3.28
#252	gpt-oss-20b	openai	1317.22	1310.81–1323.64	10,393	Apache 2.0	10.72
#253	gemma-3n-e4b-it	google	1317.17	1311.98–1322.37	22,082	Gemma	7.01
#254	qwen2.5-plus-1127	alibaba	1314.62	1308.26–1320.98	10,187	Proprietary	10.52
#255	nvidia-nemotron-3-nano-30b-a3b-bf16	nvidia	1314.24	1308.70–1319.78	15,325	NVIDIA Open Model	7.99
#256	mistral-large-2407	mistral	1314.23	1310.29–1318.16	45,459	Mistral Research	4.03
#257	athene-v2-chat	未标注	1314.10	1309.54–1318.67	24,739	NexusFlow	5.43
#258	gpt-4-0125-preview	openai	1313.00	1308.91–1317.08	93,439	Proprietary	4.35
#259	gpt-4-1106-preview	openai	1312.61	1308.71–1316.50	100,105	Proprietary	3.95
#260	hunyuan-standard-2025-02-10	tencent	1311.17	1301.50–1320.84	3,904	Proprietary	24.34
#261	gemini-1.5-flash-002	google	1309.06	1304.83–1313.29	34,902	Proprietary	4.66
#262	grok-2-mini-2024-08-13	xai	1308.30	1304.57–1312.03	52,567	Proprietary	3.63
#263	mercury	inception-ai	1307.98	1294.12–1321.84	1,930	Proprietary	49.99
#264	deepseek-v2.5	deepseek	1307.09	1302.38–1311.81	24,572	DeepSeek	5.79
#265	olmo-3-32b-think	allenai	1306.72	1298.41–1315.03	5,682	Apache 2.0	17.98
#266	athene-70b-0725	未标注	1306.39	1300.71–1312.06	19,621	CC-BY-NC-4.0	8.39
#267	granite-4.1-8b	ibm	1305.95	1295.77–1316.13	4,023	Apache 2.0	27.00
#268	mistral-large-2411	mistral	1305.42	1300.96–1309.87	28,073	MRL	5.16
#269	magistral-medium-2506	mistral	1304.70	1298.24–1311.17	11,468	Proprietary	10.87
#270	gemma-3-4b-it	google	1303.14	1293.82–1312.46	4,171	Gemma	22.61
#271	mistral-small-3.1-24b-instruct-2503	mistral	1303.07	1298.52–1307.62	32,618	Apache 2.0	5.39
#272	qwen2.5-72b-instruct	alibaba	1302.78	1298.69–1306.87	39,406	Qwen	4.35
#273	llama-3.1-nemotron-70b-instruct	nvidia	1298.67	1290.91–1306.44	7,140	Llama 3.1	15.69
#274	hunyuan-large-vision	tencent	1294.09	1284.93–1303.25	5,259	Proprietary	21.84
#275	granite-4.2-3b	ibm	1293.74	1281.87–1305.61	3,072	Apache 2.0	36.68
#276	llama-3.1-70b-instruct	meta	1293.14	1289.39–1296.89	55,240	Llama 3.1 Community	3.66
#277	amazon-nova-pro-v1.0	amazon	1290.03	1285.44–1294.61	24,745	Proprietary	5.47
#278	jamba-1.5-large	未标注	1289.28	1281.97–1296.59	8,662	Jamba Open	13.90
#279	gemma-2-27b-it	google	1289.20	1285.80–1292.60	75,754	Gemma license	3.00
#280	reka-core-20240904	未标注	1287.85	1280.66–1295.03	7,312	Proprietary	13.43
#281	gpt-4-0314	openai	1287.48	1282.71–1292.24	54,173	Proprietary	5.91
#282	granite-4.2-8b	ibm	1286.65	1275.19–1298.10	3,087	Apache 2.0	34.15
#283	llama-3.1-nemotron-51b-instruct	nvidia	1286.33	1276.36–1296.30	3,749	Llama 3.1	25.87
#284	gemini-1.5-flash-001	google	1286.31	1281.80–1290.81	62,833	Proprietary	5.28
#285	llama-3.1-tulu-3-70b	allenai	1285.92	1275.46–1296.38	2,846	Llama 3.1	28.49
#286	ibm-granite-h-small	ibm	1285.54	1277.05–1294.04	5,532	Apache 2.0	18.79
#287	olmo-3.1-32b-think	allenai	1285.50	1278.18–1292.83	8,117	Apache 2.0	13.97
#288	claude-3-sonnet-20240229	anthropic	1281.07	1277.08–1285.06	109,284	Proprietary	4.14
#289	gemma-2-9b-it-simpo	未标注	1280.10	1273.23–1286.96	10,072	MIT	12.25
#290	nemotron-4-340b-instruct	nvidia	1276.89	1271.56–1282.21	19,659	NVIDIA Open Model	7.38
#291	llama-3-70b-instruct	meta	1276.28	1272.69–1279.87	156,876	Llama 3 Community	3.36
#292	command-r-plus-08-2024	cohere	1276.07	1269.49–1282.65	9,866	CC-BY-NC-4.0	11.27
#293	gpt-4-0613	openai	1275.82	1271.76–1279.88	88,723	Proprietary	4.29
#294	mistral-small-24b-instruct-2501	mistral	1274.25	1268.33–1280.17	14,681	Apache 2.0	9.12
#295	glm-4-0520	zai	1273.20	1266.22–1280.19	9,788	Proprietary	12.70
#296	reka-flash-20240904	未标注	1272.13	1265.13–1279.13	7,536	Proprietary	12.76
#297	qwen2.5-coder-32b-instruct	alibaba	1270.47	1262.34–1278.60	5,432	Apache 2.0	17.21
#298	c4ai-aya-expanse-32b	cohere	1267.02	1262.12–1271.92	27,124	CC-BY-NC-4.0	6.25
#299	gemma-2-9b-it	google	1266.73	1262.90–1270.56	54,611	Gemma license	3.82
#300	deepseek-coder-v2	deepseek	1265.07	1258.80–1271.35	15,147	DeepSeek License	10.26
#301	qwen2-72b-instruct	alibaba	1261.53	1256.58–1266.48	37,325	Qianwen LICENSE	6.38
#302	command-r-plus	cohere	1261.51	1257.18–1265.83	77,554	CC-BY-NC-4.0	4.86
#303	claude-3-haiku-20240307	anthropic	1261.39	1257.63–1265.16	117,701	Proprietary	3.69
#304	amazon-nova-lite-v1.0	amazon	1260.33	1255.14–1265.52	19,372	Proprietary	7.02
#305	gemini-1.5-flash-8b-001	google	1258.56	1254.25–1262.86	35,558	Proprietary	4.83
#306	phi-4	microsoft	1256.13	1251.52–1260.74	24,126	MIT	5.53
#307	olmo-2-0325-32b-instruct	allenai	1251.47	1240.69–1262.25	3,334	Apache-2.0	30.25
#308	command-r-08-2024	cohere	1250.15	1243.54–1256.75	10,140	CC-BY-NC-4.0	11.36
#309	mistral-large-2402	mistral	1242.24	1237.52–1246.95	62,436	Proprietary	5.79
#310	amazon-nova-micro-v1.0	amazon	1240.59	1235.41–1245.77	19,364	Proprietary	6.98
#311	jamba-1.5-mini	未标注	1239.58	1232.35–1246.81	8,858	Jamba Open	13.60
#312	ministral-8b-2410	mistral	1237.46	1228.37–1246.55	4,781	MRL	21.51
#313	gemini-pro-dev-api	google	1236.45	1229.14–1243.77	18,354	Proprietary	13.93
#314	qwen1.5-110b-chat	alibaba	1233.89	1228.34–1239.44	26,195	Qianwen LICENSE	8.02
#315	reka-flash-21b-20240226-online	未标注	1233.28	1225.89–1240.68	15,450	Proprietary	14.24
#316	qwen1.5-72b-chat	alibaba	1233.22	1227.95–1238.50	39,302	Qianwen LICENSE	7.23
#317	hunyuan-standard-256k	tencent	1233.16	1221.43–1244.89	2,728	Proprietary	35.82
#318	mixtral-8x22b-instruct-v0.1	mistral	1229.28	1224.72–1233.84	51,416	Apache 2.0	5.41
#319	command-r	cohere	1226.58	1221.80–1231.36	54,036	CC-BY-NC-4.0	5.95
#320	reka-flash-21b-20240226	未标注	1226.55	1220.60–1232.49	24,806	Proprietary	9.19
#321	gpt-3.5-turbo-0125	openai	1225.15	1220.45–1229.85	66,207	Proprietary	5.74
#322	llama-3-8b-instruct	meta	1223.36	1219.62–1227.10	104,642	Llama 3 Community	3.64
#323	gemini-pro	google	1223.23	1211.50–1234.96	6,390	Proprietary	35.81
#324	c4ai-aya-expanse-8b	cohere	1222.87	1215.89–1229.85	9,818	CC-BY-NC-4.0	12.68
#325	mistral-medium	mistral	1222.44	1216.95–1227.93	34,550	Proprietary	7.84
#326	llama-3.1-tulu-3-8b	allenai	1220.33	1209.66–1231.01	2,896	Llama 3.1	29.66
#327	zephyr-orpo-141b-A35b-v0.1	未标注	1212.72	1201.94–1223.50	4,652	Apache 2.0	30.25
#328	yi-1.5-34b-chat	未标注	1212.65	1207.61–1217.69	24,146	Apache-2.0	6.61
#329	llama-3.1-8b-instruct	meta	1211.12	1206.95–1215.28	49,605	Llama 3.1 Community	4.51
#330	granite-3.1-8b-instruct	ibm	1208.24	1197.14–1219.33	3,090	Apache 2.0	32.05
#331	gpt-3.5-turbo-1106	openai	1203.76	1195.03–1212.49	16,619	Proprietary	19.84
#332	qwen1.5-32b-chat	alibaba	1203.69	1197.58–1209.80	21,741	Qianwen LICENSE	9.73
#333	gemma-2-2b-it	google	1199.88	1195.77–1203.98	46,616	Gemma license	4.39
#334	phi-3-medium-4k-instruct	microsoft	1197.64	1192.47–1202.80	25,055	MIT	6.94
#335	mixtral-8x7b-instruct-v0.1	mistral	1196.86	1192.60–1201.13	73,503	Apache 2.0	4.74
#336	dbrx-instruct-preview	未标注	1195.18	1189.07–1201.29	32,191	DBRX LICENSE	9.72
#337	qwen1.5-14b-chat	alibaba	1190.79	1183.68–1197.90	17,839	Qianwen LICENSE	13.17
#338	internlm2_5-20b-chat	未标注	1190.57	1183.38–1197.75	9,901	Other	13.44
#339	deepseek-llm-67b-chat	deepseek	1184.59	1173.12–1196.06	4,932	DeepSeek License	34.24
#340	wizardlm-70b	microsoft	1184.46	1175.03–1193.88	8,214	Llama 2 Community	23.13
#341	yi-34b-chat	未标注	1183.47	1176.65–1190.29	15,483	Yi License	12.11
#342	granite-3.0-8b-instruct	ibm	1182.78	1174.11–1191.46	6,638	Apache 2.0	19.60
#343	openchat-3.5	未标注	1182.68	1173.03–1192.33	7,968	Apache-2.0	24.23
#344	openchat-3.5-0106	未标注	1182.42	1174.43–1190.40	12,637	Apache-2.0	16.60
#345	gemma-1.1-7b-it	google	1182.21	1176.14–1188.28	23,893	Gemma license	9.59
#346	snowflake-arctic-instruct	未标注	1179.71	1173.80–1185.62	32,832	Apache 2.0	9.09
#347	granite-3.1-2b-instruct	ibm	1178.63	1167.44–1189.81	3,188	Apache 2.0	32.56
#348	tulu-2-dpo-70b	未标注	1177.40	1167.60–1187.19	6,535	AI2 ImpACT Low-risk	24.98
#349	openhermes-2.5-mistral-7b	未标注	1175.60	1165.21–1186.00	5,006	Apache-2.0	28.12
#350	vicuna-33b	未标注	1172.64	1166.44–1178.83	22,479	Non-commercial	9.98
#351	starling-lm-7b-beta	未标注	1170.80	1163.41–1178.20	16,056	Apache-2.0	14.25
#352	phi-3-small-8k-instruct	microsoft	1170.80	1164.88–1176.71	17,766	MIT	9.10
#353	llama-2-70b-chat	meta	1170.40	1164.91–1175.89	38,492	Llama 2 Community	7.85
#354	starling-lm-7b-alpha	未标注	1167.15	1159.09–1175.20	10,224	CC-BY-NC-4.0	16.89
#355	llama-3.2-3b-instruct	meta	1166.51	1158.81–1174.22	7,936	Llama 3.2	15.45
#356	nous-hermes-2-mixtral-8x7b-dpo	未标注	1164.16	1152.26–1176.06	3,777	Apache-2.0	36.85
#357	granite-3.0-2b-instruct	ibm	1156.34	1147.95–1164.72	6,837	Apache 2.0	18.30
#358	llama2-70b-steerlm-chat	nvidia	1154.41	1141.83–1166.99	3,585	Llama 2 Community	41.20
#359	qwq-32b-preview	alibaba	1154.18	1142.69–1165.66	3,231	Apache 2.0	34.32
#360	solar-10.7b-instruct-v1.0	未标注	1152.11	1138.97–1165.26	4,155	CC-BY-NC-4.0	44.98
#361	dolphin-2.2.1-mistral-7b	未标注	1151.96	1136.67–1167.26	1,679	Apache-2.0	60.89
#362	mpt-30b-chat	未标注	1150.40	1138.18–1162.62	2,572	CC-BY-NC-SA-4.0	38.87
#363	wizardlm-13b	microsoft	1149.21	1140.01–1158.42	7,044	Llama 2 Community	22.06
#364	mistral-7b-instruct-v0.2	mistral	1149.12	1142.47–1155.76	19,402	Apache-2.0	11.49
#365	falcon-180b-chat	未标注	1147.73	1130.71–1164.75	1,295	Falcon-180B TII License	75.43
#366	qwen1.5-7b-chat	alibaba	1143.58	1133.70–1153.46	4,737	Qianwen LICENSE	25.40
#367	phi-3-mini-4k-instruct-june-2024	microsoft	1142.84	1136.40–1149.28	12,297	MIT	10.79
#368	vicuna-13b	未标注	1141.20	1134.52–1147.89	19,367	Llama 2 Community	11.64
#369	llama-2-13b-chat	meta	1141.18	1134.46–1147.89	19,174	Llama 2 Community	11.74
#370	qwen-14b-chat	alibaba	1138.97	1128.05–1149.90	4,964	Qianwen LICENSE	31.08
#371	palm-2	google	1138.71	1129.36–1148.05	8,554	Proprietary	22.74
#372	gemma-7b-it	google	1137.37	1127.90–1146.85	8,925	Gemma license	23.37
#373	codellama-34b-instruct	meta	1136.56	1127.68–1145.44	7,366	Llama 2 Community	20.52
#374	zephyr-7b-beta	未标注	1130.53	1121.71–1139.34	11,118	MIT	20.23
#375	phi-3-mini-128k-instruct	microsoft	1129.50	1122.08–1136.92	20,685	MIT	14.33
#376	phi-3-mini-4k-instruct	microsoft	1127.86	1121.46–1134.25	20,118	MIT	10.64
#377	guanaco-33b	未标注	1127.16	1115.09–1139.23	2,921	Non-commercial	37.90
#378	zephyr-7b-alpha	未标注	1126.63	1111.07–1142.20	1,785	MIT	63.07
#379	stripedhyena-nous-7b	未标注	1121.25	1110.29–1132.22	5,182	Apache 2.0	31.28
#380	codellama-70b-instruct	meta	1118.88	1100.78–1136.99	1,143	Llama 2 Community	85.29
#381	gemma-1.1-2b-it	google	1116.28	1108.56–1124.00	10,854	Gemma license	15.52
#382	vicuna-7b	未标注	1114.95	1105.84–1124.07	6,923	Llama 2 Community	21.63
#383	smollm2-1.7b-instruct	未标注	1114.38	1100.03–1128.74	2,199	Apache 2.0	53.61
#384	llama-3.2-1b-instruct	meta	1110.85	1102.96–1118.73	8,045	Llama 3.2	16.18
#385	mistral-7b-instruct	mistral	1109.99	1100.72–1119.26	8,977	Apache 2.0	22.36
#386	llama-2-7b-chat	meta	1107.70	1100.65–1114.74	14,148	Llama 2 Community	12.92
#387	gemma-2b-it	google	1093.27	1081.80–1104.73	4,780	Gemma license	34.22
#388	qwen1.5-4b-chat	alibaba	1090.73	1081.41–1100.05	7,597	Qianwen LICENSE	22.63
#389	olmo-7b-instruct	allenai	1073.41	1062.27–1084.55	6,328	Apache-2.0	32.31
#390	koala-13b	未标注	1070.59	1060.75–1080.42	6,965	Non-commercial	25.19
#391	alpaca-13b	未标注	1069.68	1058.27–1081.09	5,745	Non-commercial	33.89
#392	gpt4all-13b-snoozy	未标注	1067.25	1052.09–1082.41	1,743	Non-commercial	59.82
#393	mpt-7b-chat	未标注	1063.00	1051.08–1074.93	3,924	CC-BY-NC-SA-4.0	37.03
#394	chatglm3-6b	未标注	1056.20	1044.56–1067.84	4,658	Apache-2.0	35.27
#395	RWKV-4-Raven-14B	未标注	1041.95	1030.61–1053.29	4,845	Apache 2.0	33.48
#396	chatglm2-6b	未标注	1024.40	1010.85–1037.95	2,658	Apache-2.0	47.81
#397	oasst-pythia-12b	未标注	1023.09	1012.27–1033.91	6,310	Apache 2.0	30.47
#398	chatglm-6b	未标注	995.49	982.98–1008.01	4,914	Non-commercial	40.76
#399	fastchat-t5-3b	未标注	992.18	979.86–1004.50	4,203	Apache 2.0	39.49
#400	dolly-v2-12b	未标注	981.97	968.51–995.43	3,412	MIT	47.16
#401	llama-13b	meta	974.64	958.93–990.34	2,391	Non-commercial	64.20
#402	stablelm-tuned-alpha-7b	未标注	952.85	940.20–965.51	3,287	CC-BY-NC-SA-4.0	41.69

TERMINAL-BENCH / 4.0

Code Agent 任务解决率

模型 × Agent × 推理档位。展示官方返回的全部 27 条记录，含同模型的不同推理档位；原榜默认视图会合并为最优档位。

来源与方法 ↗

抓取 2026年9月24日 05:19 UTC来源更新 2026年9月21日 21:28 UTCAccuracy：任务解决率；± 为 95% 置信区间

搜索 Agent / 模型 / 机构推理档位排序

27 / 27 条记录

表格可横向滚动；点击“展开”查看通过次数、Pass@k 和 Token 细分。

Terminal-Bench 4.0 全部已展示记录
原榜	模型	Agent	档位	解决率 ±95% CI	任务	发布日期	Agent 机构	模型机构	总 Token	总成本	更多指标
#1	GPT-6 Astra	Codex	max	58.18% ±2.79%	330	2026-09-03	OpenAI	OpenAI	1,529,778,322	$3,267.18	展开通过任务 192 Pass@2 / 3 / 4 / 5 64.85% / 67.88% / 69.70% / 71.21% 输出 Token 23,988,992 缓存输入 Token 1,443,351,702 未缓存输入 Token 1,505,789,330 平均任务耗时 2,796.3 秒
#2	Fable 5.1	Claude Code	max	57.88% ±3.76%	330	2026-09-01	Anthropic	Anthropic	2,746,221,560	$6,243.5	展开通过任务 191 Pass@2 / 3 / 4 / 5 70.00% / 74.70% / 77.27% / 78.79% 输出 Token 63,121,568 缓存输入 Token 2,467,056,906 未缓存输入 Token 2,683,099,992 平均任务耗时 3,893.5 秒
#2	GPT-6 Astra	Codex	xhigh	57.88% ±2.72%	330	2026-09-03	OpenAI	OpenAI	1,201,907,334	$2,350.51	展开通过任务 191 Pass@2 / 3 / 4 / 5 64.24% / 66.97% / 68.48% / 69.70% 输出 Token 14,950,287 缓存输入 Token 1,140,730,194 未缓存输入 Token 1,186,957,047 平均任务耗时 2,208.5 秒
#2	GPT-6 Astra	Codex	high	57.88% ±2.97%	330	2026-09-03	OpenAI	OpenAI	1,208,224,034	$2,269.42	展开通过任务 191 Pass@2 / 3 / 4 / 5 65.45% / 68.64% / 70.30% / 71.21% 输出 Token 13,703,943 缓存输入 Token 1,151,219,647 未缓存输入 Token 1,194,520,091 平均任务耗时 2,113.7 秒
#2	Fable 5.1	Claude Code	xhigh	57.88% ±3.36%	330	2026-09-01	Anthropic	Anthropic	2,343,414,000	$4,872.04	展开通过任务 191 Pass@2 / 3 / 4 / 5 67.58% / 71.52% / 73.94% / 75.76% 输出 Token 50,426,931 缓存输入 Token 2,140,665,130 未缓存输入 Token 2,292,987,069 平均任务耗时 3,202.1 秒
#6	Fable 5.1	Claude Code	high	54.55% ±3.44%	330	2026-09-01	Anthropic	Anthropic	2,233,621,484	$3,985.38	展开通过任务 180 Pass@2 / 3 / 4 / 5 64.70% / 69.70% / 73.03% / 75.76% 输出 Token 37,429,044 缓存输入 Token 2,064,726,577 未缓存输入 Token 2,196,192,440 平均任务耗时 3,017.7 秒
#7	GPT-6 Astra	Codex	medium	54.24% ±2.66%	330	2026-09-03	OpenAI	OpenAI	1,051,885,207	$1,914.8	展开通过任务 179 Pass@2 / 3 / 4 / 5 60.30% / 63.18% / 65.15% / 66.67% 输出 Token 10,770,605 缓存输入 Token 1,003,874,758 未缓存输入 Token 1,041,114,602 平均任务耗时 1,885.9 秒
#8	Fable 5.1	Claude Code	medium	53.94% ±3.39%	330	2026-09-01	Anthropic	Anthropic	1,620,460,859	$2,832.9	展开通过任务 178 Pass@2 / 3 / 4 / 5 63.79% / 67.88% / 70.00% / 71.21% 输出 Token 25,991,800 缓存输入 Token 1,498,585,169 未缓存输入 Token 1,594,469,059 平均任务耗时 2,528.3 秒
#8	Opus 5	Claude Code	xhigh	53.94% ±3.17%	330	2026-07-24	Anthropic	Anthropic	6,897,141,993	$6,086.22	展开通过任务 178 Pass@2 / 3 / 4 / 5 62.58% / 66.21% / 68.48% / 69.70% 输出 Token 59,162,574 缓存输入 Token 6,631,366,566 未缓存输入 Token 6,837,979,419 平均任务耗时 4,514.7 秒
#10	Opus 5	Claude Code	max	51.82% ±3.39%	330	2026-07-24	Anthropic	Anthropic	6,527,147,637	$5,969.11	展开通过任务 171 Pass@2 / 3 / 4 / 5 61.67% / 65.76% / 68.18% / 69.70% 输出 Token 66,012,156 缓存输入 Token 6,271,909,956 未缓存输入 Token 6,461,135,481 平均任务耗时 4,792.8 秒
#11	GPT-6 Astra	Codex	low	50.61% ±2.75%	330	2026-09-03	OpenAI	OpenAI	889,780,710	$1,557.3	展开通过任务 167 Pass@2 / 3 / 4 / 5 57.12% / 60.15% / 62.12% / 63.64% 输出 Token 7,988,337 缓存输入 Token 851,115,630 未缓存输入 Token 881,792,373 平均任务耗时 1,683.7 秒
#12	Opus 5	Claude Code	high	50.30% ±3.73%	330	2026-07-24	Anthropic	Anthropic	5,482,033,334	$4,662.27	展开通过任务 166 Pass@2 / 3 / 4 / 5 62.27% / 68.03% / 71.82% / 74.24% 输出 Token 47,599,774 缓存输入 Token 5,303,103,991 未缓存输入 Token 5,434,433,560 平均任务耗时 3,822.2 秒
#13	Opus 5	Claude Code	medium	44.85% ±3.83%	330	2026-07-24	Anthropic	Anthropic	3,668,060,107	$3,191.59	展开通过任务 148 Pass@2 / 3 / 4 / 5 57.42% / 63.79% / 67.27% / 69.70% 输出 Token 33,478,510 缓存输入 Token 3,541,133,339 未缓存输入 Token 3,634,581,597 平均任务耗时 3,292.9 秒
#14	Fable 5	Claude Code	max	44.55% ±3.85%	330	2026-06-09	Anthropic	Anthropic	3,785,220,369	$7,265.01	展开通过任务 147 Pass@2 / 3 / 4 / 5 57.27% / 63.18% / 66.36% / 68.18% 输出 Token 58,619,905 缓存输入 Token 3,564,065,093 未缓存输入 Token 3,726,600,464 平均任务耗时 4,202.8 秒
#15	Fable 5.1	Claude Code	low	43.33% ±3.61%	330	2026-09-01	Anthropic	Anthropic	1,334,012,516	$2,358.72	展开通过任务 143 Pass@2 / 3 / 4 / 5 54.55% / 60.00% / 63.03% / 65.15% 输出 Token 19,935,011 缓存输入 Token 1,225,657,226 未缓存输入 Token 1,314,077,505 平均任务耗时 2,290.8 秒
#16	GLM-5.3	Claude Code	max	41.82% ±3.23%	330	2026-08-14	Anthropic	Z.ai	8,677,039,373	$2,727.63	展开通过任务 138 Pass@2 / 3 / 4 / 5 50.76% / 54.70% / 56.67% / 57.58% 输出 Token 68,660,047 缓存输入 Token 8,444,041,600 未缓存输入 Token 8,608,379,326 平均任务耗时 5,831 秒
#17	Grok 4.7	Grok Build	xhigh	37.58% ±3.54%	330	2026-09-21	xAI	xAI	5,485,862,389	$3,683.287	展开通过任务 124 Pass@2 / 3 / 4 / 5 48.33% / 53.94% / 57.58% / 60.61% 输出 Token 22,977,067 缓存输入 Token 5,207,370,880 未缓存输入 Token 255,514,442 平均任务耗时 5,700.9 秒
#18	GPT-5.6 Sol	Codex	max	37.27% ±3.78%	330	2026-06-26	OpenAI	OpenAI	4,409,948,969	$2,541.7	展开通过任务 123 Pass@2 / 3 / 4 / 5 49.55% / 55.00% / 58.18% / 60.61% 输出 Token 23,451,415 缓存输入 Token 4,316,454,466 未缓存输入 Token 4,386,497,554 平均任务耗时 2,388.4 秒
#19	Opus 5	Claude Code	low	34.85% ±3.94%	330	2026-07-24	Anthropic	Anthropic	2,721,178,403	$2,393.88	展开通过任务 115 Pass@2 / 3 / 4 / 5 48.18% / 54.70% / 58.79% / 62.12% 输出 Token 23,835,411 缓存输入 Token 2,619,174,445 未缓存输入 Token 2,697,342,992 平均任务耗时 2,947.3 秒
#20	Opus 4.8	Claude Code	max	23.64% ±3.56%	330	2026-05-28	Anthropic	Anthropic	6,415,823,126	$6,481.26	展开通过任务 78 Pass@2 / 3 / 4 / 5 34.55% / 40.45% / 44.24% / 46.97% 输出 Token 88,813,309 缓存输入 Token 6,136,275,149 未缓存输入 Token 6,327,009,817 平均任务耗时 5,163 秒
#21	GPT-5.6 Terra	Codex	max	21.52% ±3.25%	330	2026-06-26	OpenAI	OpenAI	5,683,805,798	$1,733.52	展开通过任务 71 Pass@2 / 3 / 4 / 5 30.61% / 36.36% / 40.61% / 43.94% 输出 Token 33,224,833 缓存输入 Token 5,560,907,776 未缓存输入 Token 5,650,580,965 平均任务耗时 2,514.6 秒
#22	Grok 4.6	Grok Build	high	20.30% ±3.09%	330	2026-08-12	xAI	xAI	4,001,923,428	$3,591.58	展开通过任务 67 Pass@2 / 3 / 4 / 5 28.48% / 33.03% / 36.36% / 39.39% 输出 Token 32,618,221 缓存输入 Token 3,877,638,528 未缓存输入 Token 3,969,305,207 平均任务耗时 2,351.5 秒
#23	Gemini 3.8 Flash	mini-SWE-agent	high	19.09% ±3.36%	330	2026-09-02	SWE-agent	Google	17,189,729,495	$1,828.77	展开通过任务 63 Pass@2 / 3 / 4 / 5 28.79% / 35.30% / 40.00% / 43.94% 输出 Token 68,056,484 缓存输入 Token 16,692,887,408 未缓存输入 Token 17,121,673,011 平均任务耗时 1,995.1 秒
#24	GPT-5.6 Luna	Codex	max	17.27% ±2.85%	330	2026-06-26	OpenAI	OpenAI	11,557,463,640	$346.67	展开通过任务 57 Pass@2 / 3 / 4 / 5 24.24% / 28.48% / 31.21% / 33.33% 输出 Token 61,282,672 缓存输入 Token 11,305,212,851 未缓存输入 Token 11,496,180,968 平均任务耗时 4,088.3 秒
#25	Grok 4.5	Grok Build	high	12.42% ±2.62%	330	2026-07-16	xAI	xAI	3,402,536,479	$2,094.11	展开通过任务 41 Pass@2 / 3 / 4 / 5 18.33% / 22.27% / 25.15% / 27.27% 输出 Token 32,389,328 缓存输入 Token 3,248,613,376 未缓存输入 Token 3,370,147,151 平均任务耗时 3,162.1 秒
#25	Sonnet 5	Claude Code	max	12.42% ±3.06%	330	2026-06-30	Anthropic	Anthropic	21,562,180,742	$9,603.86	展开通过任务 41 Pass@2 / 3 / 4 / 5 20.45% / 26.21% / 30.91% / 34.85% 输出 Token 116,598,866 缓存输入 Token 21,027,162,464 未缓存输入 Token 21,445,581,876 平均任务耗时 6,510.3 秒
#27	Gemini 3.7 Flash	mini-SWE-agent	high	11.21% ±2.45%	330	2026-08-13	SWE-agent	Google	11,137,189,802	$1,261.87	展开通过任务 37 Pass@2 / 3 / 4 / 5 16.36% / 20.00% / 23.03% / 25.76% 输出 Token 52,134,001 缓存输入 Token 10,736,925,351 未缓存输入 Token 11,085,055,801 平均任务耗时 1,688.9 秒

GITHUB / DAILY · ALL LANGUAGES

GitHub Trending

全语言日榜，按原站页面顺序展示抓取时页面上的全部 17 条仓库；不是只筛出 Agent 工具。“今日 Star”是热度，不是质量评分。

查看来源 ↗

抓取 2026年9月24日 05:19 UTC日榜会变化；原站无按日期回看接口

搜索仓库 / 描述语言排序

17 / 17 条记录

表格可横向滚动，查看全部字段与完整仓库描述。

GitHub 全语言日榜抓取时展示的全部仓库
原榜	仓库与描述	语言	今日 Star	总 Star	Fork	贡献者
#1	anthropics/financial-services	Python	+664	37,050	5,388	@mihilmy · @cxyback · @manar-ant · @tobinsouth · @viveknair
#2	google/ax Google's open agentic orchestration runtime	Go	+1,543	9,320	448	@rakyll · @joycel-github · @wjjclaud · @anj-s · @zbl94
#3	davila7/claude-code-templates CLI tool for configuring and monitoring Claude Code	Python	+389	31,575	3,586	@davila7 · @claude · @github-actions · @claude · @dependabot
#4	BuilderIO/agent-native A framework for building agentic apps	TypeScript	+87	6,629	594	@steve8708 · @builder-io-integration · @github-actions · @builderio-bot · @3mdistal
#5	obra/superpowers An agentic skills framework & software development methodology that works.	Shell	+474	290,796	26,023	@obra · @arittr · @claude · @clkao · @gaurav0107
#6	dream-num/univer The Office Harness for AI Agents — Spreadsheets, Docs, Slides, Canvas, Relational Tables, and PDF in one runtime.	TypeScript	+1,142	16,474	1,454	@jikkai · @DR-Univer · @wzhudev · @Dushusir · @wpxp123456
#7	Open-Dev-Society/OpenStock OpenStock is an open-source alternative to expensive market platforms. Track real-time prices, set personalized alerts, and explore detailed company insights — built openly, for everyone, forever free.	TypeScript	+344	18,934	2,304	@ravixalgorithm · @coderabbitai · @keshav-005 · @nb-src · @chinnsenn
#8	agent-substrate/substrate Agent Substrate: the core system	Go	+558	3,568	419	@BenTheElder · @juli4n · @HavenXia · @thockin · @zoez7
#9	strands-agents/harness-sdk Build an agent harness and control it end-to-end. Open-source SDK for production AI agents in Python & TypeScript - any model, any cloud.	Python	+115	7,939	1,210	@zastrowm · @pgrayy · @dependabot · @strands-agent · @yonib05
#10	HKUDS/CLI-Anything "CLI-Anything: Making ALL Software Agent-Native" -- CLI-Hub: https://clianything.cc/	Python	+57	50,016	4,603	@yuh-yang · @claude · @AiMiDi · @omerarslan0 · @hito0512
#11	superdesigndev/treg OpenRouter for agent tools. Join community here: https://discord.gg/6mQYYfFMAn	Python	+506	2,812	244	@JayZeeDesign · @claude · @stonexer · @unclecode · @shehjad-dev
#12	pbakaus/impeccable The design language that makes your AI harness better at design.	JavaScript	+304	70,453	4,277	@pbakaus · @claude · @github-actions · @abdulwahabone · @cursoragent
#13	mvt-project/mvt MVT (Mobile Verification Toolkit) helps with conducting forensics of mobile devices in order to find signs of a potential compromise.	Python	+543	14,520	1,384	@botherder · @DonnchaC · @Te-k · @besendorf · @dependabot
#14	DeusData/codebase-memory-mcp High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.	C	+190	44,667	3,644	@DeusData · @claude · @shanemccarron-maker · @dependabot · @CaptainMittens
#15	harry7557558/spirula-studio Cross-vendor 3D Gaussian Splatting trainer - video to splat to mesh, Vulkan or CUDA.	C++	+69	786	57	@harry7557558 · @MotivaCG · @debsahu · @tutudoudou · @ev5unleash
#16	browser-use/video-use Edit videos with coding agents	Python	+746	26,601	3,170	@gregpr07 · @claude · @ShawnPana · @sidorovanthon · @antoinersx
#17	TNT-Likely/PanWatch 盯盘侠 PanWatch · 自托管 AI 盯盘助手，集成 TradingAgents 多 Agent 投资决策 \| A股/港股/美股实时监控、持仓管理、智能分析、全渠道推送	Python	+95	1,614	304	@TNT-Likely · @ContantLearn · @andyaa423