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: 2026-01-01 08:08:37 UTC / 2026-01-01 16:08:37 CST (北京时间)
排行榜
1
1◄─►2
gemini-3-pro
1490
±5
21,938
Proprietary
2
1◄─►6
gemini-3-flash
1480Preliminary
±8
7,020
Proprietary
3
2◄─►6
grok-4.1-thinking
1477
±5
22,618
xAI
Proprietary
4
2◄─►7
Anthropicclaude-opus-4-5-20251101-thinking-32k
1470
±6
14,992
Anthropic
Proprietary
5
2◄─►8
Anthropicclaude-opus-4-5-20251101
1467
±6
15,914
Anthropic
Proprietary
6
4◄─►8
grok-4.1
1466
±5
24,682
xAI
Proprietary
7
2◄─►9
gemini-3-flash (thinking-minimal)
1464Preliminary
±9
5,631
Proprietary
8
5◄─►12
gpt-5.1-high
1458
±5
19,416
OpenAI
Proprietary
9
8◄─►17
gemini-2.5-pro
1451
±3
82,231
Proprietary
10
8◄─►17
Anthropicclaude-sonnet-4-5-20250929-thinking-32k
1450
±4
33,533
Anthropic
Proprietary
11
9◄─►18
Anthropicclaude-opus-4-1-20250805-thinking-16k
1448
±4
49,310
Anthropic
Proprietary
12
7◄─►21
ernie-5.0-preview-1203
1448Preliminary
±9
4,697
Baidu
Proprietary
13
9◄─►20
Anthropicclaude-sonnet-4-5-20250929
1446
±5
29,093
Anthropic
Proprietary
14
8◄─►26
gpt-5.2
1443Preliminary
±12
2,466
OpenAI
Proprietary
15
9◄─►23
gpt-4.5-preview-2025-02-27
1443
±6
14,646
OpenAI
Proprietary
16
11◄─►21
Anthropicclaude-opus-4-1-20250805
1442
±3
62,378
Anthropic
Proprietary
17
12◄─►22
chatgpt-4o-latest-20250326
1441
±3
69,532
OpenAI
Proprietary
18
9◄─►26
gpt-5.2-high
1439Preliminary
±9
6,225
OpenAI
Proprietary
19
9◄─►28
glm-4.7
1439Preliminary
±9
4,557
Z.ai
MIT
20
12◄─►26
gpt-5.1
1438
±5
20,928
OpenAI
Proprietary
21
13◄─►26
gpt-5-high
1436
±5
32,735
OpenAI
Proprietary
22
16◄─►27
o3-2025-04-16
1434
±4
61,343
OpenAI
Proprietary
23
15◄─►29
qwen3-max-preview
1434
±5
28,017
Alibaba
Proprietary
24
17◄─►36
grok-4-1-fast-reasoning
1431
±6
15,352
xAI
Proprietary
25
17◄─►38
MoonshotAIkimi-k2-thinking-turbo
1430
±5
21,089
Moonshot
Modified MIT
26
17◄─►43
ernie-5.0-preview-1103
1429
±7
8,860
Baidu
Proprietary
27
23◄─►45
gpt-5-chat
1425
±4
32,013
OpenAI
Proprietary
28
24◄─►45
glm-4.6
1425
±4
29,344
Z.ai
MIT
29
21◄─►45
qwen3-max-2025-09-23
1424
±6
9,212
Alibaba
Proprietary
30
22◄─►45
deepseek-v3.2-exp
1423
±7
11,915
DeepSeek AI
MIT
31
24◄─►45
Anthropicclaude-opus-4-20250514-thinking-16k
1423
±4
37,682
Anthropic
Proprietary
32
24◄─►48
deepseek-v3.2-exp-thinking
1422
±7
9,181
DeepSeek AI
MIT
33
25◄─►45
qwen3-235b-a22b-instruct-2507
1421
±3
56,921
Alibaba
Apache 2.0
34
24◄─►50
grok-4-fast-chat
1421
±8
7,030
xAI
Proprietary
35
24◄─►49
deepseek-v3.2
1420
±6
13,952
DeepSeek AI
MIT
36
24◄─►51
deepseek-v3.2-thinking
1419
±7
11,086
DeepSeek AI
MIT
37
26◄─►50
deepseek-r1-0528
1419
±6
19,152
DeepSeek
MIT
38
25◄─►51
MoonshotAIkimi-k2-0905-preview
1418
±7
11,942
Moonshot
Modified MIT
39
26◄─►52
deepseek-v3.1
1417
±6
15,203
DeepSeek
MIT
40
27◄─►51
MoonshotAIkimi-k2-0711-preview
1416
±5
28,504
Moonshot
Modified MIT
41
26◄─►52
deepseek-v3.1-thinking
1416
±7
11,939
DeepSeek
MIT
42
24◄─►55
deepseek-v3.1-terminus
1416
±10
3,733
DeepSeek AI
MIT
43
26◄─►53
mistral-large-3
1415
±6
11,789
Mistral
Apache 2.0
44
27◄─►53
qwen3-vl-235b-a22b-instruct
1414
±7
10,249
Alibaba
Apache 2.0
45
26◄─►60
deepseek-v3.1-terminus-thinking
1414
±10
3,505
DeepSeek AI
MIT
46
33◄─►53
gpt-4.1-2025-04-14
1412
±4
52,332
OpenAI
Proprietary
47
33◄─►53
mistral-medium-2508
1412
±4
50,801
Mistral
Proprietary
48
33◄─►53
Anthropicclaude-opus-4-20250514
1412
±4
45,430
Anthropic
Proprietary
49
34◄─►55
grok-3-preview-02-24
1410
±4
34,035
xAI
Proprietary
50
37◄─►59
grok-4-0709
1409
±4
42,369
xAI
Proprietary
51
35◄─►62
glm-4.5
1409
±5
24,705
Z.ai
MIT
52
40◄─►59
gemini-2.5-flash
1408
±3
81,715
Proprietary
53
42◄─►65
gemini-2.5-flash-preview-09-2025
1405
±4
33,367
Proprietary
54
47◄─►65
grok-4-fast-reasoning
1402
±5
18,789
xAI
Proprietary
55
49◄─►65
Anthropicclaude-haiku-4-5-20251001
1402
±4
32,078
Anthropic
Proprietary
56
49◄─►66
o1-2024-12-17
1401
±4
28,039
OpenAI
Proprietary
57
49◄─►68
qwen3-next-80b-a3b-instruct
1401
±5
22,982
Alibaba
Apache 2.0
58
47◄─►70
longcat-flash-chat
1400
±6
11,455
Meituan
MIT
59
51◄─►68
qwen3-235b-a22b-no-thinking
1400
±4
39,212
Alibaba
Apache 2.0
60
53◄─►69
Anthropicclaude-sonnet-4-20250514-thinking-32k
1399
±4
36,040
Anthropic
Proprietary
61
52◄─►74
qwen3-235b-a22b-thinking-2507
1397
±6
9,304
Alibaba
Apache 2.0
62
53◄─►73
deepseek-r1
1396
±5
18,722
DeepSeek
MIT
63
49◄─►78
mimo-v2-flash (non-thinking)
1396
±10
3,882
Xiaomi
MIT
64
53◄─►76
qwen3-vl-235b-a22b-thinking
1394
±7
7,956
Alibaba
Apache 2.0
65
57◄─►75
deepseek-v3-0324
1393
±4
46,590
DeepSeek
MIT
66
56◄─►76
gpt-5-mini-high
1392
±5
27,312
OpenAI
Proprietary
67
52◄─►81
Tencenthunyuan-vision-1.5-thinking
1391
±12
2,202
Tencent
Proprietary
68
59◄─►77
o4-mini-2025-04-16
1391
±4
46,671
OpenAI
Proprietary
69
57◄─►78
mai-1-preview
1391
±5
18,095
Microsoft AI
Proprietary
70
61◄─►79
Anthropicclaude-sonnet-4-20250514
1389
±4
41,456
Anthropic
Proprietary
71
61◄─►79
o1-preview
1387
±5
31,505
OpenAI
Proprietary
72
62◄─►79
Anthropicclaude-3-7-sonnet-20250219-thinking-32k
1387
±4
39,799
Anthropic
Proprietary
73
61◄─►80
qwen3-coder-480b-a35b-instruct
1387
±5
25,017
Alibaba
Apache 2.0
74
60◄─►83
Tencenthunyuan-t1-20250711
1385
±9
4,800
Tencent
Proprietary
75
64◄─►82
mistral-medium-2505
1384
±5
34,376
Mistral
Proprietary
76
66◄─►82
qwen3-30b-a3b-instruct-2507
1382
±5
24,088
Alibaba
Apache 2.0
77
69◄─►83
gpt-4.1-mini-2025-04-14
1381
±4
40,348
OpenAI
Proprietary
78
67◄─►86
Tencenthunyuan-turbos-20250416
1381
±6
11,084
Tencent
Proprietary
79
63◄─►89
Minimaxminimax-m2.1-preview
1381Preliminary
±10
3,414
MiniMax
MIT
80
72◄─►86
gemini-2.5-flash-lite-preview-09-2025-no-thinking
1378
±4
34,924
Proprietary
81
73◄─►89
gemini-2.5-flash-lite-preview-06-17-thinking
1375
±4
33,787
Proprietary
82
74◄─►89
qwen3-235b-a22b
1374
±5
27,051
Alibaba
Apache 2.0
83
76◄─►89
qwen2.5-max
1373
±4
33,488
Alibaba
Proprietary
84
78◄─►89
Anthropicclaude-3-5-sonnet-20241022
1372
±3
89,711
Anthropic
Proprietary
85
78◄─►92
Anthropicclaude-3-7-sonnet-20250219
1371
±4
44,461
Anthropic
Proprietary
86
78◄─►92
glm-4.5-air
1371
±4
31,489
Z.ai
MIT
87
80◄─►95
qwen3-next-80b-a3b-thinking
1368
±6
13,757
Alibaba
Apache 2.0
88
80◄─►95
Minimaxminimax-m1
1366
±4
36,686
MiniMax
Apache 2.0
89
85◄─►96
gemma-3-27b-it
1365
±4
49,142
Gemma
90
80◄─►106
amazon-nova-experimental-chat-11-10
1363Preliminary
±9
4,927
Amazon
Proprietary
91
85◄─►100
o3-mini-high
1363
±5
18,725
OpenAI
Proprietary
92
85◄─►102
grok-3-mini-high
1362
±5
17,478
xAI
Proprietary
93
87◄─►103
gemini-2.0-flash-001
1360
±4
44,996
Proprietary
94
87◄─►111
deepseek-v3
1358
±5
21,995
DeepSeek
DeepSeek
95
89◄─►111
grok-3-mini-beta
1357
±5
23,674
xAI
Proprietary
96
90◄─►116
mistral-small-2506
1355
±5
18,233
Mistral
Apache 2.0
97
87◄─►117
intellect-3
1355
±8
5,435
Prime Intellect
MIT
98
91◄─►116
gpt-oss-120b
1353
±4
31,097
OpenAI
Apache 2.0
99
92◄─►117
gemini-2.0-flash-lite-preview-02-05
1353
±4
25,222
Proprietary
100
90◄─►119
glm-4.5v
1353
±8
4,963
Z.ai
MIT
101
93◄─►116
Coherecommand-a-03-2025
1352
±3
57,602
Cohere
CC-BY-NC-4.0
102
93◄─►117
gemini-1.5-pro-002
1351
±3
56,014
Proprietary
103
96◄─►119
o3-mini
1348
±3
58,670
OpenAI
Proprietary
104
94◄─►120
amazon-nova-experimental-chat-10-20
1348
±6
11,625
Amazon
Proprietary
105
90◄─►129
Tencenthunyuan-turbos-20250226
1347
±12
2,254
Tencent
Proprietary
106
94◄─►122
ling-flash-2.0
1346
±7
7,111
Ant Group
MIT
107
91◄─►129
amazon-nova-experimental-chat-10-09
1346
±11
2,877
Amazon
Proprietary
108
94◄─►124
Stepfunstep-3
1346
±7
6,605
StepFun
Apache 2.0
109
94◄─►126
Minimaxminimax-m2
1346
±8
7,078
MiniMax
Apache 2.0
110
90◄─►130
Nvidiallama-3.1-nemotron-ultra-253b-v1
1346
±12
2,573
Nvidia
Nvidia Open Model
111
99◄─►120
gpt-4o-2024-05-13
1345
±3
113,568
OpenAI
Proprietary
112
93◄─►129
qwen3-32b
1345
±9
3,944
Alibaba
Apache 2.0
113
94◄─►129
qwen-plus-0125
1345
±8
5,862
Alibaba
Proprietary
114
96◄─►130
glm-4-plus-0111
1343
±8
5,808
Zhipu
Proprietary
115
102◄─►124
Anthropicclaude-3-5-sonnet-20240620
1342
±3
82,864
Anthropic
Proprietary
116
96◄─►132
gemma-3-12b-it
1340
±9
3,866
Gemma
117
96◄─►134
Nvidianvidia-llama-3.3-nemotron-super-49b-v1.5
1340
±10
3,466
Nvidia
Nvidia Open
118
96◄─►136
Tencenthunyuan-turbo-0110
1339
±11
2,324
Tencent
Proprietary
119
102◄─►131
gpt-5-nano-high
1338
±7
8,354
OpenAI
Proprietary
120
104◄─►134
nova-2-lite
1336
±6
11,307
Amazon
Proprietary
121
106◄─►131
o1-mini
1336
±4
52,303
OpenAI
Proprietary
122
107◄─►131
Metallama-3.1-405b-instruct-bf16
1335
±4
41,948
Meta
Llama 3.1 Community
123
109◄─►134
gpt-4o-2024-08-06
1335
±4
45,787
OpenAI
Proprietary
124
109◄─►134
qwq-32b
1334
±4
26,183
Alibaba
Apache 2.0
125
110◄─►134
grok-2-2024-08-13
1334
±4
63,725
xAI
Proprietary
126
107◄─►134
gemini-advanced-0514
1334
±5
50,652
Proprietary
127
110◄─►134
Metallama-3.1-405b-instruct-fp8
1333
±3
60,273
Meta
Llama 3.1 Community
128
106◄─►145
Stepfunstep-2-16k-exp-202412
1333
±9
4,896
StepFun
Proprietary
129
116◄─►146
01.AIyi-lightning
1328
±5
27,624
01 AI
Proprietary
130
119◄─►147
Metallama-4-maverick-17b-128e-instruct
1327
±4
41,053
Meta
Llama 4
131
120◄─►149
qwen3-30b-a3b
1326
±5
27,389
Alibaba
Apache 2.0
132
110◄─►157
Nvidiallama-3.3-nemotron-49b-super-v1
1326
±12
2,246
Nvidia
Nvidia
133
114◄─►156
Tencenthunyuan-large-2025-02-10
1325
±10
3,760
Tencent
Proprietary
134
127◄─►151
gpt-4-turbo-2024-04-09
1324
±4
98,965
OpenAI
Proprietary
135
128◄─►152
Anthropicclaude-3-5-haiku-20241022
1323
±3
71,221
Anthropic
Proprietary
136
120◄─►156
deepseek-v2.5-1210
1322
±8
6,877
DeepSeek
DeepSeek
137
128◄─►153
Metallama-4-scout-17b-16e-instruct
1322
±5
31,041
Meta
Llama
138
128◄─►152
gemini-1.5-pro-001
1322
±4
79,769
Proprietary
139
128◄─►152
Anthropicclaude-3-opus-20240229
1322
±3
196,365
Anthropic
Proprietary
140
127◄─►157
gpt-4.1-nano-2025-04-14
1321
±8
6,146
OpenAI
Proprietary
141
128◄─►157
Stepfunstep-1o-turbo-202506
1320
±7
9,631
StepFun
Proprietary
142
128◄─►157
ring-flash-2.0
1320
±7
7,239
Ant Group
MIT
143
131◄─►156
Metallama-3.3-70b-instruct
1319
±3
55,946
Meta
Llama-3.3
144
129◄─►157
gemma-3n-e4b-it
1319
±5
23,408
Gemma
145
130◄─►157
glm-4-plus
1318
±5
26,343
Zhipu AI
Proprietary
146
128◄─►158
gpt-oss-20b
1318
±6
10,826
OpenAI
Apache 2.0
147
128◄─►163
Nvidianvidia-nemotron-3-nano-30b-a3b-bf16
1318Preliminary
±8
6,611
Nvidia
NVIDIA Open Model
148
131◄─►159
qwen-max-0919
1317
±6
16,599
Alibaba
Qwen
149
132◄─►157
gpt-4o-mini-2024-07-18
1317
±3
69,291
OpenAI
Proprietary
150
132◄─►165
qwen2.5-plus-1127
1314
±6
10,254
Alibaba
Proprietary
151
137◄─►163
mistral-large-2407
1313
±4
45,968
Mistral
Mistral Research
152
136◄─►164
athene-v2-chat
1313
±4
24,880
NexusFlow
NexusFlow
153
137◄─►163
gpt-4-0125-preview
1313
±4
94,534
OpenAI
Proprietary
154
137◄─►163
gpt-4-1106-preview
1313
±4
101,116
OpenAI
Proprietary
155
128◄─►168
mercury
1312
±14
1,961
Inception AI
Proprietary
156
132◄─►168
Tencenthunyuan-standard-2025-02-10
1310
±10
3,923
Tencent
Proprietary
157
140◄─►166
gemini-1.5-flash-002
1310
±4
35,180
Proprietary
158
149◄─►167
grok-2-mini-2024-08-13
1307
±4
52,792
xAI
Proprietary
159
149◄─►168
deepseek-v2.5
1306
±5
24,839
DeepSeek
DeepSeek
160
149◄─►168
athene-70b-0725
1305
±6
19,795
NexusFlow
CC-BY-NC-4.0
161
153◄─►168
mistral-large-2411
1305
±4
28,454
Mistral
MRL
162
149◄─►168
magistral-medium-2506
1305
±6
11,952
Mistral
Proprietary
163
147◄─►170
olmo-3-32b-think
1304
±8
5,986
Allen AI
Apache 2.0
164
154◄─►168
mistral-small-3.1-24b-instruct-2503
1304
±4
34,003
Mistral
Apache 2.0
165
148◄─►174
gemma-3-4b-it
1302
±9
4,196
Gemma
166
155◄─►168
qwen2.5-72b-instruct
1302
±4
39,634
Alibaba
Qwen
167
156◄─►177
Nvidiallama-3.1-nemotron-70b-instruct
1298
±8
7,216
Nvidia
Llama 3.1
168
157◄─►179
Tencenthunyuan-large-vision
1294
±9
5,571
Tencent
Proprietary
169
165◄─►177
Metallama-3.1-70b-instruct
1293
±4
56,002
Meta
Llama 3.1 Community
170
166◄─►182
jamba-1.5-large
1289
±7
8,730
AI21 Labs
Jamba Open
171
167◄─►180
amazon-nova-pro-v1.0
1288
±4
25,218
Amazon
Proprietary
172
165◄─►186
ibm-granite-h-small
1288
±8
5,682
IBM
Apache 2.0
173
167◄─►180
gemma-2-27b-it
1287
±3
76,196
Gemma license
174
166◄─►182
reka-core-20240904
1287
±7
7,380
Reka AI
Proprietary
175
167◄─►182
gpt-4-0314
1287
±5
54,754
OpenAI
Proprietary
176
166◄─►188
Nvidiallama-3.1-nemotron-51b-instruct
1286
±10
3,777
Nvidia
Llama 3.1
177
165◄─►188
llama-3.1-tulu-3-70b
1286
±10
2,882
Ai2
Llama 3.1
178
169◄─►182
gemini-1.5-flash-001
1285
±4
63,418
Proprietary
179
170◄─►188
Anthropicclaude-3-sonnet-20240229
1281
±4
110,173
Anthropic
Proprietary
180
169◄─►188
gemma-2-9b-it-simpo
1279
±7
10,108
Princeton
MIT
181
172◄─►188
Nvidianemotron-4-340b-instruct
1278
±5
19,913
Nvidia
NVIDIA Open Model
182
172◄─►189
Coherecommand-r-plus-08-2024
1277
±7
9,932
Cohere
CC-BY-NC-4.0
183
176◄─►188
Metallama-3-70b-instruct
1276
±4
158,907
Meta
Llama 3 Community
184
176◄─►189
gpt-4-0613
1275
±4
89,612
OpenAI
Proprietary
185
176◄─►191
mistral-small-24b-instruct-2501
1274
±6
14,831
Mistral
Apache 2.0
186
176◄─►193
glm-4-0520
1273
±7
9,857
Zhipu AI
Proprietary
187
176◄─►193
reka-flash-20240904
1273
±7
7,585
Reka AI
Proprietary
188
177◄─►197
qwen2.5-coder-32b-instruct
1269
±8
5,452
Alibaba
Apache 2.0
189
183◄─►197
Coherec4ai-aya-expanse-32b
1267
±5
27,362
Cohere
CC-BY-NC-4.0
190
185◄─►197
gemma-2-9b-it
1265
±4
54,957
Gemma license
191
185◄─►199
deepseek-coder-v2
1264
±6
15,242
DeepSeek AI
DeepSeek License
192
186◄─►198
Coherecommand-r-plus
1263
±4
78,400
Cohere
CC-BY-NC-4.0
193
186◄─►199
qwen2-72b-instruct
1262
±5
37,688
Alibaba
Qianwen LICENSE
194
188◄─►199
Anthropicclaude-3-haiku-20240307
1261
±4
118,626
Anthropic
Proprietary
195
188◄─►199
amazon-nova-lite-v1.0
1260
±5
19,765
Amazon
Proprietary
196
188◄─►199
gemini-1.5-flash-8b-001
1259
±4
35,914
Proprietary
197
191◄─►199
Azurephi-4
1255
±5
24,355
Microsoft
MIT
198
188◄─►205
olmo-2-0325-32b-instruct
1252
±11
3,379
Allen AI
Apache-2.0
199
192◄─►204
Coherecommand-r-08-2024
1251
±7
10,229
Cohere
CC-BY-NC-4.0
200
198◄─►208
mistral-large-2402
1242
±5
63,404
Mistral
Proprietary
201
198◄─►208
amazon-nova-micro-v1.0
1241
±5
19,776
Amazon
Proprietary
202
198◄─►213
jamba-1.5-mini
1239
±7
8,918
AI21 Labs
Jamba Open
203
198◄─►216
ministral-8b-2410
1237
±9
4,833
Mistral
MRL
204
199◄─►216
gemini-pro-dev-api
1235
±7
18,454
Proprietary
205
200◄─►216
qwen1.5-110b-chat
1234
±5
26,679
Alibaba
Qianwen LICENSE
206
200◄─►216
qwen1.5-72b-chat
1234
±5
39,689
Alibaba
Qianwen LICENSE
207
200◄─►217
reka-flash-21b-20240226-online
1233
±7
15,606
Reka AI
Proprietary
208
198◄─►218
Tencenthunyuan-standard-256k
1233
±12
2,761
Tencent
Proprietary
209
202◄─►217
mixtral-8x22b-instruct-v0.1
1230
±4
52,214
Mistral
Apache 2.0
210
202◄─►218
Coherecommand-r
1228
±5
54,710
Cohere
CC-BY-NC-4.0
211
202◄─►218
reka-flash-21b-20240226
1227
±6
25,026
Reka AI
Proprietary
212
203◄─►219
gpt-3.5-turbo-0125
1224
±5
67,214
OpenAI
Proprietary
213
203◄─►220
mistral-medium
1224
±5
34,893
Mistral
Proprietary
214
203◄─►220
Coherec4ai-aya-expanse-8b
1223
±7
9,922
Cohere
CC-BY-NC-4.0
215
207◄─►219
Metallama-3-8b-instruct
1223
±4
106,055
Meta
Llama 3 Community
216
202◄─►223
gemini-pro
1222
±12
6,418
Proprietary
217
202◄─►222
llama-3.1-tulu-3-8b
1222
±11
2,943
Ai2
Llama 3.1
218
209◄─►224
HuggingFacezephyr-orpo-141b-A35b-v0.1
1214
±11
4,712
HuggingFace
Apache 2.0
219
214◄─►223
01.AIyi-1.5-34b-chat
1213
±5
24,417
01 AI
Apache-2.0
220
216◄─►223
Metallama-3.1-8b-instruct
1211
±4
50,235
Meta
Llama 3.1 Community
221
212◄─►229
granite-3.1-8b-instruct
1210
±11
3,142
IBM
Apache 2.0
222
216◄─►228
qwen1.5-32b-chat
1205
±6
22,068
Alibaba
Qianwen LICENSE
223
217◄─►231
gpt-3.5-turbo-1106
1202
±9
16,760
OpenAI
Proprietary
224
221◄─►231
gemma-2-2b-it
1198
±4
46,900
Gemma license
225
220◄─►231
Azurephi-3-medium-4k-instruct
1198
±5
25,301
Microsoft
MIT
226
221◄─►231
mixtral-8x7b-instruct-v0.1
1198
±4
74,303
Mistral
Apache 2.0
227
221◄─►236
dbrx-instruct-preview
1196
±6
32,760
Databricks
DBRX LICENSE
228
221◄─►240
qwen1.5-14b-chat
1193
±7
18,066
Alibaba
Qianwen LICENSE
229
222◄─►240
InternLMinternlm2_5-20b-chat
1192
±7
10,038
InternLM
Other
230
223◄─►246
Azurewizardlm-70b
1185
±9
8,270
Microsoft
Llama 2 Community
231
223◄─►247
deepseek-llm-67b-chat
1184
±12
4,950
DeepSeek AI
DeepSeek License
232
227◄─►244
01.AIyi-34b-chat
1184
±7
15,624
01 AI
Yi License
233
227◄─►246
granite-3.0-8b-instruct
1184
±9
6,727
IBM
Apache 2.0
234
227◄─►246
OpenChatopenchat-3.5-0106
1183
±8
12,712
OpenChat
Apache-2.0
235
227◄─►247
OpenChatopenchat-3.5
1182
±10
8,009
OpenChat
Apache-2.0
236
227◄─►248
granite-3.1-2b-instruct
1181
±11
3,236
IBM
Apache 2.0
237
228◄─►246
Snowflakesnowflake-arctic-instruct
1180
±6
33,272
Snowflake
Apache 2.0
238
228◄─►247
gemma-1.1-7b-it
1180
±6
24,327
Gemma license
239
228◄─►249
tulu-2-dpo-70b
1179
±10
6,579
AllenAI/UW
AI2 ImpACT Low-risk
240
228◄─►251
openhermes-2.5-mistral-7b
1176
±10
5,026
NousResearch
Apache-2.0
241
230◄─►250
vicuna-33b
1174
±6
22,613
LMSYS
Non-commercial
242
230◄─►251
starling-lm-7b-beta
1173
±7
16,190
Nexusflow
Apache-2.0
243
230◄─►251
Azurephi-3-small-8k-instruct
1172
±6
17,983
Microsoft
MIT
244
231◄─►251
Metallama-2-70b-chat
1172
±5
38,767
Meta
Llama 2 Community
245
231◄─►254
starling-lm-7b-alpha
1168
±8
10,267
UC Berkeley
CC-BY-NC-4.0
246
235◄─►255
Metallama-3.2-3b-instruct
1167
±8
8,043
Meta
Llama 3.2
247
230◄─►256
nous-hermes-2-mixtral-8x7b-dpo
1166
±12
3,792
NousResearch
Apache-2.0
248
238◄─►261
qwq-32b-preview
1159
±11
3,256
Alibaba
Apache 2.0
249
239◄─►264
Nvidiallama2-70b-steerlm-chat
1157
±13
3,605
Nvidia
Llama 2 Community
250
245◄─►261
granite-3.0-2b-instruct
1157
±8
6,922
IBM
Apache 2.0
251
241◄─►265
solar-10.7b-instruct-v1.0
1154
±13
4,187
Upstage AI
CC-BY-NC-4.0
252
240◄─►269
dolphin-2.2.1-mistral-7b
1152
±15
1,685
Cognitive Computations
Apache-2.0
253
245◄─►267
mpt-30b-chat
1151
±12
2,606
MosaicML
CC-BY-NC-SA-4.0
254
247◄─►265
mistral-7b-instruct-v0.2
1151
±7
19,603
Mistral
Apache-2.0
255
246◄─►265
Azurewizardlm-13b
1150
±9
7,122
Microsoft
Llama 2 Community
256
245◄─►272
falcon-180b-chat
1148
±17
1,312
TII
Falcon-180B TII License
257
248◄─►270
qwen1.5-7b-chat
1144
±10
4,782
Alibaba
Qianwen LICENSE
258
248◄─►269
Azurephi-3-mini-4k-instruct-june-2024
1143
±6
12,414
Microsoft
MIT
259
248◄─►269
Metallama-2-13b-chat
1143
±7
19,357
Meta
Llama 2 Community
260
248◄─►270
vicuna-13b
1142
±7
19,539
LMSYS
Llama 2 Community
261
248◄─►272
qwen-14b-chat
1139
±11
5,004
Alibaba
Qianwen LICENSE
262
250◄─►272
palm-2
1137
±9
8,634
Proprietary
263
250◄─►272
Metacodellama-34b-instruct
1137
±9
7,417
Meta
Llama 2 Community
264
250◄─►272
gemma-7b-it
1135
±9
9,034
Gemma license
265
254◄─►273
HuggingFacezephyr-7b-beta
1132
±9
11,220
HuggingFace
MIT
266
255◄─►273
Azurephi-3-mini-128k-instruct
1131
±7
21,024
Microsoft
MIT
267
258◄─►273
Azurephi-3-mini-4k-instruct
1129
±6
20,539
Microsoft
MIT
268
251◄─►277
HuggingFacezephyr-7b-alpha
1129
±16
1,803
HuggingFace
MIT
269
254◄─►276
guanaco-33b
1128
±12
2,955
UW
Non-commercial
270
260◄─►277
stripedhyena-nous-7b
1121
±11
5,214
Together AI
Apache 2.0
271
255◄─►278
Metacodellama-70b-instruct
1119
±18
1,151
Meta
Llama 2 Community
272
260◄─►277
HuggingFacesmollm2-1.7b-instruct
1119
±14
2,244
HuggingFace
Apache 2.0
273
265◄─►277
vicuna-7b
1116
±9
6,972
LMSYS
Llama 2 Community
274
268◄─►277
gemma-1.1-2b-it
1114
±8
11,035
Gemma license
275
268◄─►277
Metallama-3.2-1b-instruct
1113
±8
8,166
Meta
Llama 3.2
276
268◄─►278
mistral-7b-instruct
1111
±9
9,042
Mistral
Apache 2.0
277
269◄─►278
Metallama-2-7b-chat
1109
±7
14,272
Meta
Llama 2 Community
278
275◄─►282
gemma-2b-it
1091
±12
4,817
Gemma license
279
278◄─►280
qwen1.5-4b-chat
1091
±9
7,662
Alibaba
Qianwen LICENSE
280
278◄─►285
olmo-7b-instruct
1075
±11
6,412
Allen AI
Apache-2.0
281
279◄─►285
koala-13b
1071
±10
6,998
UC Berkeley
Non-commercial
282
280◄─►285
alpaca-13b
1067
±11
5,828
Stanford
Non-commercial
283
279◄─►286
gpt4all-13b-snoozy
1066
±15
1,773
Nomic AI
Non-commercial
284
280◄─►286
mpt-7b-chat
1062
±12
3,977
MosaicML
CC-BY-NC-SA-4.0
285
280◄─►286
chatglm3-6b
1057
±12
4,692
Tsinghua
Apache-2.0
286
283◄─►288
RWKVRWKV-4-Raven-14B
1042
±11
4,898
RWKV
Apache 2.0
287
286◄─►288
chatglm2-6b
1026
±14
2,683
Tsinghua
Apache-2.0
288
286◄─►288
oasst-pythia-12b
1023
±11
6,343
OpenAssistant
Apache 2.0
289
289◄─►292
chatglm-6b
996
±13
4,968
Tsinghua
Non-commercial
290
289◄─►292
fastchat-t5-3b
992
±12
4,270
LMSYS
Apache 2.0
291
289◄─►292
dolly-v2-12b
980
±14
3,471
Databricks
MIT
292
289◄─►293
Metallama-13b
972
±16
2,441
Meta
Non-commercial
293
292◄─►293
Stabilitystablelm-tuned-alpha-7b
953
±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。
最后更新于
这有帮助吗？