模型榜单

这是一个基于 Chatbot Arena (lmarena.ai) 数据的排行榜，通过自动化流程生成。

数据更新时间: 2025-10-23 08:08:10 UTC / 2025-10-23 16:08:10 CST (北京时间)

排行榜

排名 (UB)	模型	分数	95% 置信区间 (±)	票数	组织/公司	许可证
1	gemini-2.5-pro 1	1451	±4	54,087	Google	Proprietary
1	claude-opus-4-1-20250805-thinking-16k 1	1447	±5	21,306	Anthropic	Proprietary
1	claude-sonnet-4-5-20250929-thinking-32k 1	1445	±8	6,287	Anthropic	Proprietary
1	gpt-4.5-preview-2025-02-27 1	1441	±6	14,644	OpenAI	Proprietary
2	chatgpt-4o-latest-20250326 1	1440	±4	40,013	OpenAI	Proprietary
2	o3-2025-04-16 1	1440	±4	51,293	OpenAI	Proprietary
2	claude-sonnet-4-5-20250929 1	1438	±8	6,144	Anthropic	Proprietary
2	gpt-5-high 1	1437	±5	23,580	OpenAI	Proprietary
2	claude-opus-4-1-20250805 1	1437	±5	33,298	Anthropic	Proprietary
3	qwen3-max-preview 1	1434	±6	18,078	Alibaba	Proprietary
10	gpt-5-chat 1	1425	±5	21,630	OpenAI	Proprietary
10	qwen3-max-2025-09-23 1	1423	±7	6,919	Alibaba	Proprietary
10	glm-4.6 1	1422	±9	4,401	Z.ai	MIT
11	grok-4-fast 1	1420	±8	7,104	xAI	Proprietary
11	claude-opus-4-20250514-thinking-16k 1	1419	±5	35,522	Anthropic	Proprietary
11	deepseek-v3.2-exp-thinking 1	1419	±9	4,320	DeepSeek AI	MIT
11	qwen3-vl-235b-a22b-instruct 1	1418	±8	6,312	Alibaba	Apache 2.0
11	qwen3-235b-a22b-instruct-2507 1	1418	±5	29,343	Alibaba	Apache 2.0
11	deepseek-r1-0528 1	1417	±6	19,284	DeepSeek	MIT
11	kimi-k2-0905-preview 1	1417	±7	10,772	Moonshot	Modified MIT
11	deepseek-v3.1 1	1416	±6	15,380	DeepSeek	MIT
11	deepseek-v3.1-thinking 1	1415	±7	12,098	DeepSeek	MIT
11	kimi-k2-0711-preview 1	1415	±5	28,321	Moonshot	Modified MIT
11	deepseek-v3.1-terminus 1	1414	±10	3,775	DeepSeek AI	MIT
11	deepseek-v3.1-terminus-thinking 1	1413	±10	3,541	DeepSeek AI	MIT
12	grok-4-0709 1	1413	±5	29,264	xAI	Proprietary
12	claude-opus-4-20250514 1	1411	±4	43,310	Anthropic	Proprietary
12	deepseek-v3.2-exp 1	1408	±9	4,684	DeepSeek AI	MIT
13	gpt-4.1-2025-04-14 1	1411	±4	41,918	OpenAI	Proprietary
14	grok-3-preview-02-24 1	1409	±4	34,154	xAI	Proprietary
18	mistral-medium-2508 1	1406	±5	23,844	Mistral	Proprietary
18	glm-4.5 1	1406	±5	22,612	Z.ai	MIT
18	gemini-2.5-flash-preview-09-2025 1	1404	±7	6,730	Google	Proprietary
23	claude-haiku-4-5-20251001 1	1397	±12	2,380	Anthropic	Proprietary
24	qwen3-next-80b-a3b-instruct 1	1402	±6	12,793	Alibaba	Apache 2.0
29	o1-2024-12-17 1	1400	±4	28,039	OpenAI	Proprietary
29	longcat-flash-chat 1	1398	±6	11,667	Meituan	MIT
29	qwen3-235b-a22b-thinking-2507 1	1397	±6	9,386	Alibaba	Apache 2.0
30	claude-sonnet-4-20250514-thinking-32k 1	1398	±5	33,827	Anthropic	Proprietary
30	qwen3-235b-a22b-no-thinking 1	1398	±5	39,528	Alibaba	Apache 2.0
32	gpt-5-mini-high 1	1395	±6	18,172	OpenAI	Proprietary
32	deepseek-r1 1	1394	±5	18,718	DeepSeek	MIT
32	qwen3-vl-235b-a22b-thinking 1	1392	±8	5,956	Alibaba	Apache 2.0
36	deepseek-v3-0324 1	1391	±4	44,482	DeepSeek	MIT
36	o4-mini-2025-04-16 1	1391	±4	41,513	OpenAI	Proprietary
36	mai-1-preview 1	1389	±6	14,528	Microsoft AI	Proprietary
38	claude-sonnet-4-20250514 1	1389	±5	39,329	Anthropic	Proprietary
38	hunyuan-t1-20250711 1	1384	±9	4,845	Tencent	Proprietary
39	o1-preview 1	1386	±5	31,505	OpenAI	Proprietary
39	qwen3-30b-a3b-instruct-2507 1	1385	±5	21,853	Alibaba	Apache 2.0
40	claude-3-7-sonnet-20250219-thinking-32k 1	1386	±4	39,987	Anthropic	Proprietary
41	qwen3-coder-480b-a35b-instruct 1	1384	±5	23,287	Alibaba	Apache 2.0
44	mistral-medium-2505 1	1381	±5	34,539	Mistral	Proprietary
44	hunyuan-turbos-20250416 1	1379	±6	11,135	Tencent	Proprietary
47	gpt-4.1-mini-2025-04-14 1	1379	±4	40,621	OpenAI	Proprietary
50	gemini-2.5-flash-lite-preview-09-2025-no-thinking 1	1374	±7	6,765	Google	Proprietary
52	gemini-2.5-flash-lite-preview-06-17-thinking 1	1374	±5	31,701	Google	Proprietary
52	qwen3-235b-a22b 1	1372	±5	27,210	Alibaba	Apache 2.0
54	qwen2.5-max 1	1372	±4	33,541	Alibaba	Proprietary
54	glm-4.5-air 1	1369	±5	21,945	Z.ai	MIT
55	Claude 3.5 Sonnet (10/22) 1	1370	±3	89,889	Anthropic	Proprietary
55	qwen3-next-80b-a3b-thinking 1	1367	±6	11,522	Alibaba	Apache 2.0
56	minimax-m1 1	1368	±5	31,897	MiniMax	Apache 2.0
59	grok-3-mini-high 1	1362	±5	17,615	xAI	Proprietary
59	o3-mini-high 1	1362	±5	18,735	OpenAI	Proprietary
60	gemma-3-27b-it 1	1363	±4	44,508	Google	Gemma
63	grok-3-mini-beta 1	1356	±5	23,839	xAI	Proprietary
63	deepseek-v3 1	1356	±5	21,994	DeepSeek	DeepSeek
64	glm-4.5v 1	1351	±8	5,028	Z.ai	MIT
65	mistral-small-2506 1	1353	±5	18,374	Mistral	Apache 2.0
67	gemini-2.0-flash-lite-preview-02-05 1	1352	±4	25,215	Google	Proprietary
67	Gemini-1.5-Pro-002 1	1351	±3	56,012	Google	Proprietary
67	command-a-03-2025 1	1349	±4	47,441	Cohere	CC-BY-NC-4.0
67	gpt-oss-120b 1	1348	±5	21,119	OpenAI	Apache 2.0
67	hunyuan-turbos-20250226 1	1345	±12	2,250	Tencent	Proprietary
67	llama-3.1-nemotron-ultra-253b-v1 1	1344	±12	2,573	Nvidia	Nvidia Open Model
67	amazon-nova-experimental-chat-10-09 1	1344Preliminary	±11	2,911	Amazon	Proprietary
67	qwen3-32b 1	1344	±9	3,943	Alibaba	Apache 2.0
68	qwen-plus-0125 1	1343	±8	5,861	Alibaba	Proprietary
69	o3-mini 1	1347	±3	58,935	OpenAI	Proprietary
69	step-3 1	1344	±7	6,686	StepFun	Apache 2.0
69	glm-4-plus-0111 1	1343	±8	5,806	Zhipu	Proprietary
69	ling-flash-2.0 1	1341Preliminary	±9	4,893	Ant Group	MIT
69	gemma-3-12b-it 1	1340	±10	3,866	Google	Gemma
69	nvidia-llama-3.3-nemotron-super-49b-v1.5 1	1339	±10	3,488	Nvidia	Nvidia Open
69	hunyuan-turbo-0110 1	1337	±11	2,322	Tencent	Proprietary
72	gpt-4o-2024-05-13 1	1344	±3	113,568	OpenAI	Proprietary
73	Claude 3.5 Sonnet (06/20) 1	1341	±3	82,864	Anthropic	Proprietary
73	gpt-5-nano-high 1	1337	±7	8,465	OpenAI	Proprietary
77	llama-3.1-405b-instruct-bf16 1	1335	±4	41,932	Meta	Llama 3.1 Community
77	step-2-16k-exp-202412 1	1331	±8	4,895	StepFun	Proprietary
78	o1-mini 1	1334	±3	52,301	OpenAI	Proprietary
78	GPT-4o (08/06) 1	1333	±4	45,787	OpenAI	Proprietary
78	gemini-advanced-0514 1	1332	±5	50,654	Google	Proprietary
78	qwq-32b 1	1332	±4	26,309	Alibaba	Apache 2.0
79	llama-3.1-405b-instruct-fp8 1	1333	±3	60,272	Meta	Llama 3.1 Community
79	grok-2-2024-08-13 1	1333	±3	63,725	xAI	Proprietary
79	llama-3.3-nemotron-49b-super-v1 1	1324	±12	2,243	Nvidia	Nvidia
83	hunyuan-large-2025-02-10 1	1323	±10	3,760	Tencent	Proprietary
86	yi-lightning 1	1327	±5	27,624	01 AI	Proprietary
87	llama-4-maverick-17b-128e-instruct 1	1327	±4	41,319	Meta	Llama 4
91	qwen3-30b-a3b 1	1325	±5	27,520	Alibaba	Apache 2.0
93	deepseek-v2.5-1210 1	1321	±8	6,877	DeepSeek	DeepSeek
96	gpt-4-turbo-2024-04-09 1	1323	±4	98,965	OpenAI	Proprietary
96	llama-4-scout-17b-16e-instruct 1	1322	±5	31,329	Meta	Llama
96	gpt-4.1-nano-2025-04-14 1	1319	±8	6,143	OpenAI	Proprietary
97	claude-3-opus-20240229 1	1321	±3	196,368	Anthropic	Proprietary
97	claude-3-5-haiku-20241022 1	1320	±3	71,507	Anthropic	Propretary
97	gemini-1.5-pro-001 1	1320	±4	79,769	Google	Proprietary
97	step-1o-turbo-202506 1	1319	±7	9,685	StepFun	Proprietary
97	gemma-3n-e4b-it 1	1318	±5	23,541	Google	Gemma
97	gpt-oss-20b 1	1317	±6	10,906	OpenAI	Apache 2.0
97	ring-flash-2.0 1	1314Preliminary	±9	4,971	Ant Group	MIT
100	llama-3.3-70b-instruct 1	1319	±3	56,024	Meta	Llama-3.3
100	glm-4-plus 1	1317	±5	26,342	Zhipu AI	Proprietary
100	qwen-max-0919 1	1316	±6	16,598	Alibaba	Qwen
101	qwen2.5-plus-1127 1	1313	±6	10,252	Alibaba	Proprietary
102	GPT-4o-mini (07/18) 1	1315	±3	69,290	OpenAI	Proprietary
102	hunyuan-standard-2025-02-10 1	1309	±10	3,920	Tencent	Proprietary
105	gpt-4-1106-preview 1	1313	±4	101,117	OpenAI	Proprietary
106	gpt-4-0125-preview 1	1313	±4	94,534	OpenAI	Proprietary
106	mistral-large-2407 1	1312	±4	45,968	Mistral	Mistral Research
106	athene-v2-chat 1	1312	±4	24,880	NexusFlow	NexusFlow
107	Gemini-1.5-Flash-002 1	1310	±4	35,180	Google	Proprietary
114	gemma-3-4b-it 1	1302	±9	4,195	Google	Gemma
117	magistral-medium-2506 1	1304	±6	12,018	Mistral	Proprietary
118	grok-2-mini-2024-08-13 1	1306	±4	52,789	xAI	Proprietary
118	deepseek-v2.5 1	1305	±5	24,839	DeepSeek	DeepSeek
118	athene-70b-0725 1	1304	±6	19,796	NexusFlow	CC-BY-NC-4.0
120	mistral-large-2411 1	1303	±4	28,455	Mistral	MRL
122	mistral-small-3.1-24b-instruct-2503 1	1301	±5	31,747	Mistral	Apache 2.0
124	qwen2.5-72b-instruct 1	1300	±4	39,632	Alibaba	Qwen
124	llama-3.1-nemotron-70b-instruct 1	1296	±8	7,216	Nvidia	Llama 3.1
125	hunyuan-large-vision 1	1293	±9	5,606	Tencent	Proprietary
132	Meta-Llama-3.1-70B-Instruct 1	1292	±4	56,003	Meta	Llama 3.1 Community
132	jamba-1.5-large 1	1287	±7	8,730	AI21 Labs	Jamba Open
132	reka-core-20240904 1	1286	±7	7,380	Reka AI	Proprietary
132	llama-3.1-tulu-3-70b 1	1285	±10	2,881	Ai2	Llama 3.1
132	llama-3.1-nemotron-51b-instruct 1	1285	±10	3,777	Nvidia	Llama 3.1
133	amazon-nova-pro-v1.0 1	1287	±4	25,218	Amazon	Proprietary
133	gpt-4-0314 1	1285	±5	54,754	OpenAI	Proprietary
134	gemma-2-27b-it 1	1285	±3	76,195	Google	Gemma license
135	gemini-1.5-flash-001 1	1283	±4	63,418	Google	Proprietary
135	claude-3-sonnet-20240229 1	1280	±4	110,173	Anthropic	Proprietary
135	gemma-2-9b-it-simpo 1	1277	±7	10,108	Princeton	MIT
136	command-r-plus-08-2024 1	1276	±6	9,931	Cohere	CC-BY-NC-4.0
137	nemotron-4-340b-instruct 1	1277	±5	19,913	Nvidia	NVIDIA Open Model
140	glm-4-0520 1	1272	±7	9,857	Zhipu AI	Proprietary
141	reka-flash-20240904 1	1272	±7	7,583	Reka AI	Proprietary
142	llama-3-70b-instruct 1	1275	±3	158,908	Meta	Llama 3 Community
142	gpt-4-0613 1	1274	±4	89,612	OpenAI	Proprietary
142	mistral-small-24b-instruct-2501 1	1272	±6	14,830	Mistral	Apache 2.0
143	qwen2.5-coder-32b-instruct 1	1268	±8	5,452	Alibaba	Apache 2.0
148	c4ai-aya-expanse-32b 1	1266	±5	27,362	Cohere	CC-BY-NC-4.0
150	command-r-plus 1	1263	±4	78,401	Cohere	CC-BY-NC-4.0
150	deepseek-coder-v2 1	1262	±6	15,242	DeepSeek AI	DeepSeek License
151	gemma-2-9b-it 1	1263	±4	54,954	Google	Gemma license
151	qwen2-72b-instruct 1	1261	±5	37,688	Alibaba	Qianwen LICENSE
153	claude-3-haiku-20240307 1	1260	±4	118,626	Anthropic	Proprietary
153	gemini-1.5-flash-8b-001 1	1259	±4	35,914	Google	Proprietary
153	amazon-nova-lite-v1.0 1	1259	±5	19,760	Amazon	Proprietary
153	olmo-2-0325-32b-instruct 1	1252	±11	3,377	Allen AI	Apache-2.0
156	phi-4 1	1254	±4	24,354	Microsoft	MIT
157	command-r-08-2024 1	1251	±6	10,229	Cohere	CC-BY-NC-4.0
163	mistral-large-2402 1	1242	±5	63,404	Mistral	Proprietary
163	amazon-nova-micro-v1.0 1	1240	±5	19,774	Amazon	Proprietary
163	jamba-1.5-mini 1	1237	±7	8,918	AI21 Labs	Jamba Open
163	ministral-8b-2410 1	1235	±9	4,833	Mistral	MRL
164	hunyuan-standard-256k 1	1232	±12	2,761	Tencent	Proprietary
165	qwen1.5-110b-chat 1	1234	±5	26,679	Alibaba	Qianwen LICENSE
165	qwen1.5-72b-chat 1	1233	±5	39,689	Alibaba	Qianwen LICENSE
165	reka-flash-21b-20240226-online 1	1233	±7	15,606	Reka AI	Proprietary
165	gemini-pro-dev-api 1	1233	±7	18,454	Google	Proprietary
167	mixtral-8x22b-instruct-v0.1 1	1229	±4	52,214	Mistral	Apache 2.0
167	command-r 1	1228	±5	54,710	Cohere	CC-BY-NC-4.0
167	reka-flash-21b-20240226 1	1226	±6	25,026	Reka AI	Proprietary
167	llama-3.1-tulu-3-8b 1	1221	±11	2,943	Ai2	Llama 3.1
167	gemini-pro 1	1220	±12	6,418	Google	Proprietary
168	mistral-medium 1	1223	±5	34,893	Mistral	Proprietary
168	c4ai-aya-expanse-8b 1	1222	±7	9,922	Cohere	CC-BY-NC-4.0
170	gpt-3.5-turbo-0125 1	1223	±5	67,214	OpenAI	Proprietary
171	llama-3-8b-instruct 1	1222	±4	106,055	Meta	Llama 3 Community
174	zephyr-orpo-141b-A35b-v0.1 1	1213	±11	4,712	HuggingFace	Apache 2.0
175	granite-3.1-8b-instruct 1	1209	±11	3,142	IBM	Apache 2.0
179	yi-1.5-34b-chat 1	1213	±5	24,417	01 AI	Apache-2.0
181	llama-3.1-8b-instruct 1	1210	±4	50,234	Meta	Llama 3.1 Community
181	qwen1.5-32b-chat 1	1205	±6	22,068	Alibaba	Qianwen LICENSE
182	gpt-3.5-turbo-1106 1	1200	±9	16,760	OpenAI	Proprietary
185	phi-3-medium-4k-instruct 1	1198	±5	25,301	Microsoft	MIT
186	mixtral-8x7b-instruct-v0.1 1	1197	±4	74,303	Mistral	Apache 2.0
186	gemma-2-2b-it 1	1197	±4	46,901	Google	Gemma license
186	dbrx-instruct-preview 1	1195	±6	32,760	Databricks	DBRX LICENSE
186	qwen1.5-14b-chat 1	1192	±7	18,066	Alibaba	Qianwen LICENSE
186	internlm2_5-20b-chat 1	1192	±7	10,038	InternLM	Other
188	wizardlm-70b 1	1185	±9	8,270	Microsoft	Llama 2 Community
188	deepseek-llm-67b-chat 1	1183	±11	4,950	DeepSeek AI	DeepSeek License
192	yi-34b-chat 1	1184	±7	15,624	01 AI	Yi License
192	granite-3.0-8b-instruct 1	1182	±9	6,727	IBM	Apache 2.0
192	openchat-3.5-0106 1	1182	±8	12,712	OpenChat	Apache-2.0
192	openchat-3.5 1	1181	±10	8,009	OpenChat	Apache-2.0
192	granite-3.1-2b-instruct 1	1180	±11	3,235	IBM	Apache 2.0
193	snowflake-arctic-instruct 1	1180	±6	33,272	Snowflake	Apache 2.0
193	tulu-2-dpo-70b 1	1179	±10	6,579	AllenAI/UW	AI2 ImpACT Low-risk
193	openhermes-2.5-mistral-7b 1	1175	±10	5,026	NousResearch	Apache-2.0
195	gemma-1.1-7b-it 1	1178	±6	24,327	Google	Gemma license
195	vicuna-33b 1	1173	±6	22,613	LMSYS	Non-commercial
195	starling-lm-7b-beta 1	1173	±7	16,190	Nexusflow	Apache-2.0
195	phi-3-small-8k-instruct 1	1172	±6	17,983	Microsoft	MIT
195	nous-hermes-2-mixtral-8x7b-dpo 1	1166	±12	3,792	NousResearch	Apache-2.0
196	llama-2-70b-chat 1	1171	±5	38,767	Meta	Llama 2 Community
196	starling-lm-7b-alpha 1	1168	±8	10,267	UC Berkeley	CC-BY-NC-4.0
198	llama-3.2-3b-instruct 1	1166	±8	8,043	Meta	Llama 3.2
201	qwq-32b-preview 1	1160	±11	3,256	Alibaba	Apache 2.0
203	llama2-70b-steerlm-chat 1	1157	±13	3,605	Nvidia	Llama 2 Community
205	dolphin-2.2.1-mistral-7b 1	1151	±15	1,685	Cognitive Computations	Apache-2.0
206	solar-10.7b-instruct-v1.0 1	1154	±13	4,187	Upstage AI	CC-BY-NC-4.0
210	granite-3.0-2b-instruct 1	1156	±8	6,922	IBM	Apache 2.0
210	mpt-30b-chat 1	1150	±12	2,606	MosaicML	CC-BY-NC-SA-4.0
210	falcon-180b-chat 1	1146	±17	1,312	TII	Falcon-180B TII License
211	wizardlm-13b 1	1149	±9	7,122	Microsoft	Llama 2 Community
212	mistral-7b-instruct-v0.2 1	1150	±7	19,603	Mistral	Apache-2.0
213	qwen1.5-7b-chat 1	1144	±10	4,782	Alibaba	Qianwen LICENSE
213	qwen-14b-chat 1	1138	±11	5,004	Alibaba	Qianwen LICENSE
214	phi-3-mini-4k-instruct-june-2024 1	1143	±6	12,415	Microsoft	MIT
214	llama-2-13b-chat 1	1142	±7	19,357	Meta	Llama 2 Community
214	vicuna-13b 1	1141	±7	19,539	LMSYS	Llama 2 Community
215	codellama-34b-instruct 1	1136	±9	7,417	Meta	Llama 2 Community
215	palm-2 1	1135	±9	8,634	Google	Proprietary
217	gemma-7b-it 1	1133	±9	9,034	Google	Gemma license
217	zephyr-7b-beta 1	1132	±9	11,220	HuggingFace	MIT
217	zephyr-7b-alpha 1	1128	±16	1,803	HuggingFace	MIT
219	guanaco-33b 1	1128	±12	2,955	UW	Non-commercial
220	phi-3-mini-128k-instruct 1	1130	±7	21,024	Microsoft	MIT
220	codellama-70b-instruct 1	1119	±18	1,151	Meta	Llama 2 Community
223	phi-3-mini-4k-instruct 1	1129	±6	20,539	Microsoft	MIT
225	stripedhyena-nous-7b 1	1120	±11	5,214	Together AI	Apache 2.0
225	smollm2-1.7b-instruct 1	1118	±14	2,244	HuggingFace	Apache 2.0
230	vicuna-7b 1	1114	±9	6,972	LMSYS	Llama 2 Community
233	llama-3.2-1b-instruct 1	1112	±8	8,166	Meta	Llama 3.2
233	gemma-1.1-2b-it 1	1112	±8	11,035	Google	Gemma license
233	mistral-7b-instruct 1	1110	±9	9,042	Mistral	Apache 2.0
234	llama-2-7b-chat 1	1108	±7	14,272	Meta	Llama 2 Community
241	gemma-2b-it 1	1089	±11	4,817	Google	Gemma license
243	qwen1.5-4b-chat 1	1090	±9	7,662	Alibaba	Qianwen LICENSE
243	olmo-7b-instruct 1	1075	±11	6,412	Allen AI	Apache-2.0
244	koala-13b 1	1070	±10	6,998	UC Berkeley	Non-commercial
244	gpt4all-13b-snoozy 1	1064	±15	1,773	Nomic AI	Non-commercial
245	alpaca-13b 1	1064	±11	5,828	Stanford	Non-commercial
245	mpt-7b-chat 1	1060	±12	3,977	MosaicML	CC-BY-NC-SA-4.0
245	chatglm3-6b 1	1056	±12	4,692	Tsinghua	Apache-2.0
248	RWKV-4-Raven-14B 1	1041	±11	4,898	RWKV	Apache 2.0
251	chatglm2-6b 1	1025	±14	2,683	Tsinghua	Apache-2.0
251	oasst-pythia-12b 1	1021	±11	6,343	OpenAssistant	Apache 2.0
254	chatglm-6b 1	995	±13	4,968	Tsinghua	Non-commercial
254	fastchat-t5-3b 1	990	±12	4,270	LMSYS	Apache 2.0
254	dolly-v2-12b 1	977	±14	3,471	Databricks	MIT
254	llama-13b 1	968	±16	2,441	Meta	Non-commercial
256	stablelm-tuned-alpha-7b 1	952	±13	3,325	Stability AI	CC-BY-NC-SA-4.0

说明

• 排名 (UB)：基于 Bradley-Terry 模型计算的排名。此排名反映了模型在竞技场中的综合表现，并提供了其 Elo 分数的 上界 估计，帮助理解模型的潜在竞争力。

• 模型：大型语言模型 (LLM) 的名称。部分模型名称可能已嵌入相关链接。

• 分数：模型在竞技场中通过用户投票获得的 Elo 评分。Elo 评分是一种相对排名系统，分数越高表示模型表现越好。

• 95% 置信区间 (±)：模型 Elo 评分的95%置信区间（例如：±6）。这个区间越小，表示模型的评分越稳定和可靠。

• 票数：该模型在竞技场中收到的总投票数量。投票数越多，通常意味着其评分的统计可靠性越高。

• 组织/公司：提供该模型的组织或公司。

• 许可证：模型的许可协议类型，例如专有 (Proprietary)、Apache 2.0、MIT 等。

数据来源与更新频率

本排行榜数据由自动化脚本直接从 ^{1 2} 官方网站获取。此排行榜由 GitHub Actions 每天自动更新。

免责声明

本报告仅供参考。排行榜数据是动态变化的，并基于特定时间段内用户在 Chatbot Arena 上的偏好投票。数据的完整性和准确性取决于上游数据源。不同模型可能采用不同的许可协议，使用时请务必参考模型提供商的官方说明。