大模型排行榜 · french

数据来自 LMArena（Chatbot Arena）：全球用户对匿名模型两两盲测投票，按 Elo 机制计算排名——相比跑分基准更接近真实使用体验。本页每周同步官方榜单，当前榜单发布于 2026-07-30。点击模型名可查看其各类别成绩与评分走势。

综合编程创意写作中文英文 exclude_ties expert french german hard_prompts hard_prompts_english industry_business_and_management_and_financial_o industry_entertainment_and_sports_and_media industry_legal_and_government industry_life_and_physical_and_social_science industry_mathematical industry_medicine_and_healthcare industry_software_and_it_services

排名	模型	开发方	Elo 评分	投票数
#1	claude-fable-5	anthropic	1538	—
#2	muse-spark	meta	1526	—
#3	gpt-5.6-terra-xhigh	openai	1525	—
#4	muse-spark-1.1	meta	1523	—
#5	claude-opus-4-8-thinking	anthropic	1516	—
#6	claude-opus-4-7-thinking	anthropic	1516	—
#7	claude-opus-4-7	anthropic	1514	—
#8	claude-opus-5-high	anthropic	1513	—
#9	claude-opus-4-6-thinking	anthropic	1513	—
#10	gemini-3-pro	google	1508	—
#11	gemini-3-flash	google	1505	—
#12	grok-4.20-beta-0309-reasoning	xai	1504	—
#13	gpt-5.5	openai	1503	—
#14	gemini-3.1-pro-preview	google	1502	—
#15	gemini-3.6-flash	google	1502	—
#16	gpt-5.4-high	openai	1501	—
#17	gpt-5.6-sol-xhigh	openai	1501	—
#18	claude-opus-4-8	anthropic	1500	—
#19	claude-opus-4-5-20251101	anthropic	1498	—
#20	claude-opus-4-6	anthropic	1495	—
#21	grok-4.20-multi-agent-beta-0309	xai	1494	—
#22	mimo-v2.5-pro	xiaomi	1493	—
#23	gpt-5.5-instant	openai	1491	—
#24	claude-sonnet-5-high	anthropic	1491	—
#25	glm-5.2-max	zai	1490	—
#26	gpt-5.5-high	openai	1489	—
#27	grok-4.20-beta1	xai	1488	—
#28	gpt-5.2-chat-latest-20260210	openai	1488	—
#29	claude-opus-4-5-20251101-thinking-32k	anthropic	1487	—
#30	grok-4.1-thinking	xai	1487	—
#31	glm-5.1	zai	1487	—
#32	gpt-5.4	openai	1486	—
#33	dola-seed-2.0-pro	bytedance	1483	—
#34	ernie-5.1	baidu	1483	—
#35	qwen3.7-plus	alibaba	1482	—
#36	gemini-3.5-flash-lite	google	1482	—
#37	claude-sonnet-4-6	anthropic	1482	—
#38	gemini-3.5-flash-medium	google	1481	—
#39	qwen3.5-max-preview	alibaba	1481	—
#40	mimo-v2-pro	xiaomi	1479	—
#41	deepseek-v4-pro	deepseek	1479	—
#42	gemma-4-26b-a4b	google	1478	—
#43	gemini-3-flash (thinking-minimal)	google	1478	—
#44	glm-5	zai	1478	—
#45	ernie-5.0-preview-1203	baidu	1477	—
#46	kimi-k2.6	moonshot	1477	—
#47	claude-sonnet-4-5-20250929	anthropic	1477	—
#48	grok-4.1	xai	1477	—
#49	gpt-5.2-high	openai	1477	—
#50	claude-sonnet-4-5-20250929-thinking-32k	anthropic	1476	—
#51	gpt-5.4-mini-high	openai	1476	—
#52	grok-4.5	xai	1475	—
#53	o3-2025-04-16	openai	1474	—
#54	claude-opus-4-1-20250805-thinking-16k	anthropic	1473	—
#55	qwen3.6-max-preview	alibaba	1472	—
#56	gpt-5.3-chat-latest	openai	1471	—
#57	deepseek-v4-pro-thinking	deepseek	1470	—
#58	gpt-5.1-high	openai	1470	—
#59	gemini-3.5-flash-high	google	1466	—
#60	ernie-5.0-0110	baidu	1465	—
#61	chatgpt-4o-latest-20250326	openai	1465	—
#62	gpt-5.2	openai	1465	—
#63	qwen3.5-397b-a17b	alibaba	1465	—
#64	kimi-k2.5-thinking	moonshot	1464	—
#65	gemini-2.5-pro	google	1463	—
#66	claude-opus-4-1-20250805	anthropic	1462	—
#67	qwen3.6-plus	alibaba	1462	—
#68	gemma-4-31b	google	1462	—
#69	qwen3-max-preview	alibaba	1461	—
#70	inkling	thinky	1461	—
#71	gpt-5.6-luna-xhigh	openai	1460	—
#72	qwen3-235b-a22b-instruct-2507	alibaba	1459	—
#73	deepseek-v4-flash-thinking	deepseek	1459	—
#74	gpt-5.1	openai	1458	—
#75	grok-4.3	xai	1458	—
#76	minimax-m3	minimax	1456	—
#77	glm-5v-turbo	zai	1456	—
#78	qwen3-vl-235b-a22b-instruct	alibaba	1456	—
#79	mimo-v2-omni	xiaomi	1454	—
#80	grok-3-preview-02-24	xai	1453	—
#81	mimo-v2.5	xiaomi	1451	—
#82	nvidia-nemotron-3-ultra-550b-a55b-nvfp4	nvidia	1451	—
#83	deepseek-v3.1	deepseek	1451	—
#84	qwen3.5-122b-a10b	alibaba	1449	—
#85	gemini-3.1-flash-lite-preview	google	1448	—
#86	deepseek-v3.2-exp	deepseek	1448	—
#87	glm-4.6	zai	1447	—
#88	claude-opus-4-20250514-thinking-16k	anthropic	1444	—
#89	gpt-5-chat	openai	1444	—
#90	deepseek-v4-flash	deepseek	1444	—
#91	gpt-4.5-preview-2025-02-27	openai	1444	—
#92	kimi-k2-thinking-turbo	moonshot	1444	—
#93	glm-4.7	zai	1443	—
#94	grok-4-fast-reasoning	xai	1442	—
#95	grok-4-1-fast-reasoning	xai	1442	—
#96	deepseek-v3.2	deepseek	1441	—
#97	mistral-large-3	mistral	1441	—
#98	gpt-5-high	openai	1440	—
#99	hunyuan-hy3-preview	tencent	1439	—
#100	gemini-2.5-flash-preview-09-2025	google	1439	—
#101	mistral-medium-3.5	mistral	1436	—
#102	mistral-medium-2508	mistral	1435	—
#103	minimax-m2.7	minimax	1435	—
#104	longcat-flash-chat	meituan	1433	—
#105	grok-4-0709	xai	1430	—
#106	gpt-5.4-nano-high	openai	1430	—
#107	trinity-large-preview	—	1429	—
#108	claude-sonnet-4-20250514	anthropic	1427	—
#109	claude-haiku-4-5-20251001	anthropic	1427	—
#110	deepseek-v3.2-thinking	deepseek	1427	—
#111	mimo-v2-flash (thinking)	xiaomi	1427	—
#112	qwen3-coder-480b-a35b-instruct	alibaba	1426	—
#113	deepseek-v3.2-exp-thinking	deepseek	1426	—
#114	longcat-flash-chat-2602-exp	meituan	1426	—
#115	gemini-2.5-flash	google	1426	—
#116	qwen3.5-35b-a3b	alibaba	1424	—
#117	deepseek-r1-0528	deepseek	1423	—
#118	claude-opus-4-20250514	anthropic	1422	—
#119	kimi-k2-0711-preview	moonshot	1422	—
#120	deepseek-v3-0324	deepseek	1420	—
#121	gpt-4.1-2025-04-14	openai	1420	—
#122	claude-sonnet-4-20250514-thinking-32k	anthropic	1419	—
#123	qwen3-30b-a3b-instruct-2507	alibaba	1418	—
#124	kimi-k2-0905-preview	moonshot	1418	—
#125	kimi-k2.5-instant	moonshot	1417	—
#126	qwen3.5-flash	alibaba	1416	—
#127	glm-4.5	zai	1415	—
#128	step-3.5-flash	stepfun	1412	—
#129	qwen3.5-27b	alibaba	1411	—
#130	minimax-m1	minimax	1411	—
#131	qwen2.5-max	alibaba	1409	—
#132	o4-mini-2025-04-16	openai	1408	—
#133	glm-4.5-air	zai	1405	—
#134	gpt-4.1-mini-2025-04-14	openai	1405	—
#135	qwen3-next-80b-a3b-instruct	alibaba	1404	—
#136	minimax-m2.1-preview	minimax	1403	—
#137	gemini-2.0-flash-001	google	1402	—
#138	gpt-5-mini-high	openai	1400	—
#139	mimo-v2-flash (non-thinking)	xiaomi	1398	—
#140	trinity-large-thinking	—	1398	—
#141	minimax-m2.5	minimax	1398	—
#142	amazon-nova-experimental-chat-11-10	amazon	1395	—
#143	nvidia-nemotron-3-super-120b-a12b	nvidia	1395	—
#144	gemini-2.5-flash-lite-preview-06-17-thinking	google	1393	—
#145	hunyuan-turbos-20250416	tencent	1392	—
#146	mistral-medium-2505	mistral	1391	—
#147	deepseek-v3.1-thinking	deepseek	1391	—
#148	gemini-2.5-flash-lite-preview-09-2025-no-thinking	google	1391	—
#149	deepseek-r1	deepseek	1390	—
#150	claude-3-5-sonnet-20241022	anthropic	1390	—
#151	o1-preview	openai	1389	—
#152	gemma-3-27b-it	google	1388	—
#153	claude-3-7-sonnet-20250219	anthropic	1388	—
#154	qwen3-235b-a22b-no-thinking	alibaba	1386	—
#155	o1-2024-12-17	openai	1385	—
#156	amazon-nova-experimental-chat-10-20	amazon	1383	—
#157	command-a-03-2025	cohere	1382	—
#158	gemini-2.0-flash-lite-preview-02-05	google	1376	—
#159	qwen-max-0919	alibaba	1375	—
#160	qwen3-235b-a22b	alibaba	1374	—
#161	mistral-large-2411	mistral	1373	—
#162	o3-mini-high	openai	1371	—
#163	deepseek-v3	deepseek	1371	—
#164	claude-3-7-sonnet-20250219-thinking-32k	anthropic	1368	—
#165	qwen3-next-80b-a3b-thinking	alibaba	1367	—
#166	gemini-advanced-0514	google	1365	—
#167	qwen3-30b-a3b	alibaba	1365	—
#168	mistral-small-2506	mistral	1363	—
#169	glm-4.7-flash	zai	1363	—
#170	claude-3-5-sonnet-20240620	anthropic	1362	—
#171	nvidia-nemotron-3-nano-30b-a3b-bf16	nvidia	1360	—
#172	gpt-4o-2024-05-13	openai	1359	—
#173	athene-v2-chat	—	1357	—
#174	grok-3-mini-beta	xai	1357	—
#175	o3-mini	openai	1356	—
#176	gpt-oss-120b	openai	1355	—
#177	olmo-3.1-32b-instruct	allenai	1352	—
#178	grok-2-2024-08-13	xai	1351	—
#179	grok-3-mini-high	xai	1348	—
#180	qwq-32b	alibaba	1347	—
#181	nova-2-lite	amazon	1346	—
#182	glm-4-plus	zai	1344	—
#183	claude-3-5-haiku-20241022	anthropic	1343	—
#184	gemini-1.5-pro-002	google	1342	—
#185	gemma-3n-e4b-it	google	1340	—
#186	yi-lightning	—	1340	—
#187	llama-3.1-405b-instruct-bf16	meta	1339	—
#188	gpt-4-1106-preview	openai	1339	—
#189	claude-3-opus-20240229	anthropic	1338	—
#190	gpt-4o-mini-2024-07-18	openai	1337	—
#191	gpt-4-0125-preview	openai	1337	—
#192	gpt-4-turbo-2024-04-09	openai	1334	—
#193	llama-4-scout-17b-16e-instruct	meta	1333	—
#194	gemini-1.5-pro-001	google	1331	—
#195	llama-3.1-405b-instruct-fp8	meta	1327	—
#196	mistral-large-2407	mistral	1327	—
#197	gpt-4o-2024-08-06	openai	1327	—
#198	magistral-medium-2506	mistral	1327	—
#199	athene-70b-0725	—	1326	—
#200	deepseek-v2.5	deepseek	1325	—

说明：Elo 评分由 LMArena 根据真实用户盲测投票计算，分差越大代表偏好差距越显著；排名会随每期投票变化。本页仅同步展示并翻译整理，原始数据版权归 LMArena 所有。