Benchmarks
Intelligence test
Models are ranked by pairwise comparison, not by an absolute score. On every prompt each pair of models is compared head-to-head, and a Bradley-Terry model fits a rating from those comparisons. The win rate below turns that rating into the question it answers: how often this model beats an average model on this board. 50% is average.
8,228
scored runs
111,690
pairwise comparisons
24
prompts
4,629
runs without a score (provider error)
22 of 24 prompts separate models at all. The rest are answered equally well by nearly everyone and carry no ranking signal — that is what an absolute score near 100 was really measuring.
- Coding 4/4
- Creative 2/2
- Factual 3/4
- Multilingual 7/8
- Reasoning 4/4
- Healthcare 2/2
Rank 1OpenAI
73%±2
Win rate
Elo 1174
gpt-5.5-2026-04-23
Tier A
- Won / lost / tied
- 1062 / 83 / 1366
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 96.9
By category
Coding80%
Creative89%
Factual68%
Multilingual67%
Reasoning74%
Rank 1OpenAI
73%±2
Win rate
Elo 1169
gpt-5.1
Tier B
- Won / lost / tied
- 1047 / 97 / 1367
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 96.3
By category
Coding75%
Creative79%
Factual71%
Multilingual70%
Reasoning79%
Rank 1OpenAI
72%±2
Win rate
Elo 1166
gpt-4.1-mini
Tier C
- Won / lost / tied
- 1190 / 178 / 1344
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 96.4
By category
Coding79%
Creative91%
Factual71%
Multilingual60%
Reasoning77%
Healthcare75%
Rank 1OpenAI
71%±2
Win rate
Elo 1157
gpt-4.1-2025-04-14
Tier C
- Won / lost / tied
- 1085 / 133 / 1494
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 96.6
By category
Coding71%
Creative92%
Factual71%
Multilingual65%
Reasoning73%
Healthcare84%
Rank 1OpenAI
70%±2
Win rate
Elo 1148
gpt-5.5
Tier C
- Won / lost / tied
- 1015 / 188 / 1308
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 95.1
By category
Coding69%
Creative82%
Factual74%
Multilingual66%
Reasoning74%
Rank 1OpenAI
70%±2
Win rate
Elo 1148
gpt-5.2-2025-12-11
Tier B
- Won / lost / tied
- 1009 / 185 / 1317
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 95.5
By category
Coding80%
Creative78%
Factual52%
Multilingual69%
Reasoning79%
Rank 1OpenAI
70%±2
Win rate
Elo 1146
gpt-4.1
Tier B
- Won / lost / tied
- 1060 / 184 / 1468
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 96.2
By category
Coding75%
Creative83%
Factual66%
Multilingual64%
Reasoning73%
Healthcare79%
Rank 1OpenAI
70%±2
Win rate
Elo 1145
gpt-5.2
Tier B
- Won / lost / tied
- 1017 / 208 / 1286
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 95.7
By category
Coding64%
Creative92%
Factual72%
Multilingual66%
Reasoning74%
Rank 2OpenAI
69%±2
Win rate
Elo 1143
gpt-5.4-mini
Tier A
- Won / lost / tied
- 1022 / 228 / 1261
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 95.6
By category
Coding72%
Creative92%
Factual59%
Multilingual63%
Reasoning79%
Rank 2Anthropic
69%±2
Win rate
Elo 1141
Claude Opus 4.6
Tier Barchived
- Won / lost / tied
- 1105 / 258 / 1349
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 97.6
By category
Coding61%
Creative93%
Factual73%
Multilingual64%
Reasoning74%
Healthcare90%
Rank 2OpenAI
69%±2
Win rate
Elo 1140
gpt-4.1-mini-2025-04-14
Tier C
- Won / lost / tied
- 1042 / 204 / 1466
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 95.6
By category
Coding72%
Creative82%
Factual64%
Multilingual68%
Reasoning74%
Healthcare70%
Rank 2Z.ai (GLM / Zhipu)
69%±2
Win rate
Elo 1137
GLM-4.7
Tier A
- Won / lost / tied
- 781 / 223 / 817
- prompts
- 14
- comparisons
- 1821
- Judge mean
- 92.7
By category
Coding60%
Creative85%
Factual75%
Reasoning74%
Rank 4OpenAI
68%±2
Win rate
Elo 1132
gpt-5.4-2026-03-05
Tier B
- Won / lost / tied
- 979 / 254 / 1278
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 96.1
By category
Coding72%
Creative89%
Factual57%
Multilingual64%
Reasoning74%
Rank 5OpenAI
68%±2
Win rate
Elo 1127
gpt-5.4
Tier A
- Won / lost / tied
- 941 / 244 / 1326
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 96.0
By category
Coding65%
Creative90%
Factual57%
Multilingual68%
Reasoning74%
Rank 5OpenAI
67%±2
Win rate
Elo 1125
gpt-5.1-2025-11-13
Tier B
- Won / lost / tied
- 903 / 223 / 1385
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 95.3
By category
Coding72%
Creative73%
Factual62%
Multilingual64%
Reasoning74%
Rank 5Z.ai (GLM / Zhipu)
67%±2
Win rate
Elo 1120
GLM-4.5
Tier A
- Won / lost / tied
- 775 / 279 / 830
- prompts
- 14
- comparisons
- 1884
- Judge mean
- 91.5
By category
Coding75%
Creative85%
Factual64%
Reasoning52%
Rank 6OpenAI
66%±2
Win rate
Elo 1119
gpt-4o-2024-05-13
Tier Carchived
- Won / lost / tied
- 979 / 281 / 1452
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 95.9
By category
Coding64%
Creative82%
Factual66%
Multilingual64%
Reasoning73%
Healthcare75%
Rank 12OpenAI
65%±2
Win rate
Elo 1110
gpt-4o
Tier C
- Won / lost / tied
- 979 / 344 / 1389
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 95.3
By category
Coding57%
Creative73%
Factual73%
Multilingual64%
Reasoning73%
Healthcare67%
Rank 12OpenAI
65%±2
Win rate
Elo 1108
gpt-5.4-mini-2026-03-17
Tier A
- Won / lost / tied
- 899 / 324 / 1288
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 94.2
By category
Coding64%
Creative89%
Factual54%
Multilingual63%
Reasoning74%
Rank 12OpenAI
65%±2
Win rate
Elo 1106
gpt-5-search-api
Tier C
- Won / lost / tied
- 971 / 361 / 1380
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 95.0
By category
Coding67%
Creative86%
Factual59%
Multilingual53%
Reasoning77%
Healthcare87%
Rank 13OpenAI
64%±2
Win rate
Elo 1103
- Won / lost / tied
- 1008 / 417 / 1287
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 94.5
By category
Coding66%
Creative82%
Factual60%
Multilingual52%
Reasoning78%
Healthcare89%
Rank 13Anthropic
64%±2
Win rate
Elo 1103
Claude Opus 4.5
Tier Barchived
- Won / lost / tied
- 976 / 386 / 1350
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 96.5
By category
Coding53%
Creative58%
Factual78%
Multilingual63%
Reasoning74%
Healthcare88%
Rank 14Anthropic
63%±2
Win rate
Elo 1096
Claude Sonnet 5
Tier Aarchived
- Won / lost / tied
- 586 / 299 / 655
- prompts
- 16
- comparisons
- 1540
- Judge mean
- 89.7
By category
Coding62%
Creative89%
Factual61%
Multilingual65%
Reasoning53%
Rank 16Anthropic
63%±2
Win rate
Elo 1091
Claude Opus 4.7
Tier Barchived
- Won / lost / tied
- 953 / 448 / 1311
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 95.5
By category
Coding47%
Creative60%
Factual81%
Multilingual60%
Reasoning73%
Healthcare81%
Rank 18Google Gemini
62%±2
Win rate
Elo 1088
Gemini Flash-Lite Latest
Tier C
- Won / lost / tied
- 990 / 504 / 1218
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 95.2
By category
Coding65%
Creative45%
Factual69%
Multilingual61%
Reasoning63%
Healthcare85%
Rank 16Z.ai (GLM / Zhipu)
62%±2
Win rate
Elo 1088
GLM-5
Tier A
- Won / lost / tied
- 552 / 255 / 731
- prompts
- 14
- comparisons
- 1538
- Judge mean
- 94.4
By category
Coding57%
Creative89%
Factual64%
Reasoning62%
Rank 17Google Gemini
62%±2
Win rate
Elo 1087
Nano Banana
Tier Barchived
- Won / lost / tied
- 615 / 274 / 910
- prompts
- 23
- comparisons
- 1799
- Judge mean
- 93.5
By category
Coding51%
Creative70%
Factual61%
Multilingual65%
Reasoning74%
Healthcare58%
Rank 18OpenAI
62%±2
Win rate
Elo 1085
gpt-4o-mini
Tier C
- Won / lost / tied
- 900 / 436 / 1376
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 90.2
By category
Coding60%
Creative72%
Factual54%
Multilingual67%
Reasoning68%
Healthcare54%
Rank 18OVH AI Endpoints (GRA)
62%±2
Win rate
Elo 1084
gpt-oss-120b
Tier C
- Won / lost / tied
- 850 / 429 / 1232
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 93.1
By category
Coding59%
Creative84%
Factual62%
Multilingual54%
Reasoning70%
Rank 18OpenAI
62%±2
Win rate
Elo 1084
gpt-4.1-nano-2025-04-14
Tier Carchived
- Won / lost / tied
- 903 / 449 / 1360
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 92.9
By category
Coding71%
Creative66%
Factual61%
Multilingual54%
Reasoning68%
Healthcare35%
Rank 18OpenAI
62%±2
Win rate
Elo 1083
o1
Tier Carchived
- Won / lost / tied
- 801 / 398 / 1186
- prompts
- 20
- comparisons
- 2385
- Judge mean
- 89.6
By category
Coding76%
Creative55%
Factual43%
Multilingual65%
Reasoning60%
Rank 18Anthropic
62%±2
Win rate
Elo 1083
Claude Sonnet 4.6
Tier Aarchived
- Won / lost / tied
- 893 / 446 / 1373
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 95.1
By category
Coding56%
Creative54%
Factual66%
Multilingual60%
Reasoning73%
Healthcare82%
Rank 19OpenAI
61%±2
Win rate
Elo 1080
o1-2024-12-17
Tier Carchived
- Won / lost / tied
- 780 / 396 / 1209
- prompts
- 20
- comparisons
- 2385
- Judge mean
- 86.0
By category
Coding76%
Creative6%
Factual60%
Multilingual68%
Reasoning60%
Rank 21Anthropic
61%±2
Win rate
Elo 1077
Claude Fable 5
Tier Aarchived
- Won / lost / tied
- 805 / 427 / 1320
- prompts
- 20
- comparisons
- 2552
- Judge mean
- 89.2
By category
Coding63%
Creative67%
Factual49%
Multilingual65%
Reasoning68%
Rank 21OpenAI
61%±2
Win rate
Elo 1076
gpt-4o-2024-11-20
Tier C
- Won / lost / tied
- 871 / 470 / 1371
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 94.8
By category
Coding53%
Creative50%
Factual71%
Multilingual58%
Reasoning72%
Healthcare86%
Rank 23OpenAI
61%±2
Win rate
Elo 1074
o4-mini
Tier Carchived
- Won / lost / tied
- 789 / 444 / 1152
- prompts
- 20
- comparisons
- 2385
- Judge mean
- 80.6
By category
Coding70%
Creative79%
Factual50%
Multilingual57%
Reasoning54%
Rank 23OpenAI
60%±2
Win rate
Elo 1074
gpt-4o-2024-08-06
Tier C
- Won / lost / tied
- 895 / 512 / 1305
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 93.2
By category
Coding52%
Creative73%
Factual68%
Multilingual62%
Reasoning60%
Healthcare72%
Rank 23OpenAI
60%±2
Win rate
Elo 1073
gpt-4-turbo-2024-04-09
Tier Carchived
- Won / lost / tied
- 902 / 524 / 1286
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 94.3
By category
Coding53%
Creative77%
Factual62%
Multilingual55%
Reasoning73%
Healthcare72%
Rank 23Anthropic
60%±2
Win rate
Elo 1073
Claude Sonnet 4.5
Tier Barchived
- Won / lost / tied
- 911 / 536 / 1265
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 94.5
By category
Coding39%
Creative57%
Factual74%
Multilingual61%
Reasoning77%
Healthcare78%
Rank 23OpenAI
60%±2
Win rate
Elo 1072
o3
Tier C
- Won / lost / tied
- 758 / 432 / 1123
- prompts
- 20
- comparisons
- 2313
- Judge mean
- 82.6
By category
Coding64%
Creative88%
Factual57%
Multilingual53%
Reasoning54%
Rank 23Google Gemini
60%±3
Win rate
Elo 1070
Nano Banana 2
Tier B
- Won / lost / tied
- 501 / 290 / 608
- prompts
- 18
- comparisons
- 1399
- Judge mean
- 92.6
By category
Coding70%
Creative61%
Factual53%
Multilingual67%
Reasoning65%
Healthcare36%
Rank 23OpenAI
60%±2
Win rate
Elo 1068
o3-mini-2025-01-31
Tier Carchived
- Won / lost / tied
- 762 / 451 / 1235
- prompts
- 20
- comparisons
- 2448
- Judge mean
- 90.4
By category
Coding68%
Creative72%
Factual42%
Multilingual64%
Reasoning57%
Rank 23OpenAI
59%±2
Win rate
Elo 1066
gpt-4-turbo
Tier Carchived
- Won / lost / tied
- 867 / 542 / 1303
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 94.1
By category
Coding48%
Creative71%
Factual70%
Multilingual56%
Reasoning65%
Healthcare78%
Rank 23OpenAI
59%±2
Win rate
Elo 1065
o4-mini-2025-04-16
Tier Barchived
- Won / lost / tied
- 722 / 437 / 1154
- prompts
- 20
- comparisons
- 2313
- Judge mean
- 80.4
By category
Coding57%
Creative69%
Factual45%
Multilingual70%
Reasoning61%
Rank 23OpenAI
59%±2
Win rate
Elo 1063
gpt-5.4-nano-2026-03-17
Tier A
- Won / lost / tied
- 788 / 505 / 1218
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 87.2
By category
Coding59%
Creative30%
Factual49%
Multilingual69%
Reasoning74%
Rank 26OpenAI
58%±2
Win rate
Elo 1059
o3-2025-04-16
Tier Barchived
- Won / lost / tied
- 648 / 414 / 1161
- prompts
- 19
- comparisons
- 2223
- Judge mean
- 81.0
By category
Coding65%
Creative84%
Factual43%
Multilingual56%
Reasoning54%
Rank 26Google Gemini
58%±2
Win rate
Elo 1058
Gemini 3.1 Flash Lite
Tier Barchived
- Won / lost / tied
- 735 / 484 / 1292
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 93.5
By category
Coding54%
Creative42%
Factual60%
Multilingual59%
Reasoning74%
Rank 26Anthropic
58%±2
Win rate
Elo 1058
Claude Opus 4.8
Tier Aarchived
- Won / lost / tied
- 778 / 529 / 1204
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 94.6
By category
Coding46%
Creative44%
Factual70%
Multilingual58%
Reasoning74%
Rank 31OpenAI
58%±2
Win rate
Elo 1055
gpt-4o-mini-2024-07-18
Tier C
- Won / lost / tied
- 862 / 611 / 1239
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 89.4
By category
Coding61%
Creative70%
Factual49%
Multilingual58%
Reasoning63%
Healthcare49%
Rank 34OpenAI
57%±2
Win rate
Elo 1051
gpt-4.1-nano
Tier Carchived
- Won / lost / tied
- 831 / 610 / 1271
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 91.7
By category
Coding57%
Creative60%
Factual66%
Multilingual53%
Reasoning64%
Healthcare21%
Rank 34OVH AI Endpoints (GRA)
57%±2
Win rate
Elo 1048
- Won / lost / tied
- 726 / 541 / 1244
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 93.0
By category
Coding45%
Creative69%
Factual63%
Multilingual51%
Reasoning74%
Rank 46Anthropic
55%±2
Win rate
Elo 1032
Claude Haiku 4.5
Tier Aarchived
- Won / lost / tied
- 807 / 720 / 1185
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 92.1
By category
Coding60%
Creative55%
Factual50%
Multilingual55%
Reasoning58%
Healthcare33%
Rank 49OVH AI Endpoints (GRA)
54%±2
Win rate
Elo 1029
- Won / lost / tied
- 737 / 682 / 1092
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 92.5
By category
Coding65%
Creative41%
Factual50%
Multilingual50%
Reasoning58%
Rank 50OVH AI Endpoints (GRA)
54%±2
Win rate
Elo 1027
- Won / lost / tied
- 784 / 741 / 986
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 90.4
By category
Coding58%
Creative36%
Factual52%
Multilingual53%
Reasoning62%
Rank 50OVH AI Endpoints (GRA)
54%±2
Win rate
Elo 1024
Qwen2.5-VL-72B-Instruct
Tier B
- Won / lost / tied
- 734 / 707 / 1070
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 92.4
By category
Coding55%
Creative65%
Factual34%
Multilingual52%
Reasoning74%
Rank 52OpenAI
52%±2
Win rate
Elo 1016
gpt-5.4-nano
Tier C
- Won / lost / tied
- 776 / 802 / 933
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 87.1
By category
Coding53%
Creative32%
Factual33%
Multilingual68%
Reasoning64%
Rank 52Google Gemini
50%±2
Win rate
Elo 1003
Gemini Robotics-ER 1.6 Preview
Tier Barchived
- Won / lost / tied
- 485 / 536 / 778
- prompts
- 23
- comparisons
- 1799
- Judge mean
- 79.3
By category
Coding44%
Creative82%
Factual43%
Multilingual55%
Reasoning43%
Healthcare43%
Rank 55Google Gemini
50%±2
Win rate
Elo 998
Gemini 2.5 Flash-Lite
Tier Barchived
- Won / lost / tied
- 766 / 925 / 1021
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 91.2
By category
Coding38%
Creative51%
Factual49%
Multilingual57%
Reasoning58%
Healthcare53%
Rank 55OpenAI
50%±2
Win rate
Elo 998
o3-mini
Tier Carchived
- Won / lost / tied
- 585 / 719 / 1081
- prompts
- 20
- comparisons
- 2385
- Judge mean
- 84.0
By category
Coding57%
Creative30%
Factual31%
Multilingual57%
Reasoning61%
Rank 55OpenAI
49%±2
Win rate
Elo 995
gpt-5-mini
Tier C
- Won / lost / tied
- 515 / 647 / 1078
- prompts
- 20
- comparisons
- 2240
- Judge mean
- 74.0
By category
Coding59%
Creative50%
Factual40%
Multilingual45%
Reasoning54%
Rank 56Anthropic
49%±2
Win rate
Elo 993
Claude Opus 5
Tier Aarchived
- Won / lost / tied
- 521 / 671 / 824
- prompts
- 18
- comparisons
- 2016
- Judge mean
- 83.0
By category
Coding26%
Creative86%
Factual48%
Multilingual41%
Reasoning61%
Rank 56OpenAI
49%±2
Win rate
Elo 992
gpt-4
Tier Carchived
- Won / lost / tied
- 778 / 979 / 955
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 88.9
By category
Coding55%
Creative33%
Factual34%
Multilingual54%
Reasoning58%
Healthcare43%
Rank 57OpenAI
48%±2
Win rate
Elo 986
gpt-4-0613
Tier Carchived
- Won / lost / tied
- 728 / 973 / 1011
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 90.8
By category
Coding47%
Creative53%
Factual38%
Multilingual48%
Reasoning62%
Healthcare51%
Rank 57OpenAI
46%±2
Win rate
Elo 973
gpt-3.5-turbo-1106
Tier Carchived
- Won / lost / tied
- 680 / 1012 / 1020
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 83.0
By category
Coding55%
Creative54%
Factual22%
Multilingual56%
Reasoning48%
Healthcare18%
Rank 60OpenAI
45%±2
Win rate
Elo 967
gpt-5-mini-2025-08-07
Tier Barchived
- Won / lost / tied
- 515 / 824 / 974
- prompts
- 20
- comparisons
- 2313
- Judge mean
- 70.7
By category
Coding63%
Creative35%
Factual40%
Multilingual32%
Reasoning54%
Rank 61OpenAI
45%±2
Win rate
Elo 964
gpt-3.5-turbo-0125
Tier Carchived
- Won / lost / tied
- 669 / 1065 / 978
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 83.5
By category
Coding57%
Creative46%
Factual20%
Multilingual51%
Reasoning51%
Healthcare25%
Rank 65OpenAI
42%±2
Win rate
Elo 944
gpt-3.5-turbo-16k
Tier Carchived
- Won / lost / tied
- 662 / 1194 / 856
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 82.0
By category
Coding62%
Creative30%
Factual23%
Multilingual47%
Reasoning35%
Healthcare20%
Rank 64Z.ai (GLM / Zhipu)
42%±3
Win rate
Elo 944
GLM-5.2
Tier A
- Won / lost / tied
- 361 / 639 / 538
- prompts
- 14
- comparisons
- 1538
- Judge mean
- 58.3
By category
Coding56%
Creative4%
Factual55%
Reasoning20%
Rank 67OpenAI
40%±2
Win rate
Elo 933
gpt-3.5-turbo
Tier Carchived
- Won / lost / tied
- 624 / 1236 / 852
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 83.1
By category
Coding45%
Creative54%
Factual20%
Multilingual43%
Reasoning49%
Healthcare22%
Rank 67OpenAI
40%±2
Win rate
Elo 932
gpt-5
Tier C
- Won / lost / tied
- 378 / 811 / 812
- prompts
- 19
- comparisons
- 2001
- Judge mean
- 60.3
By category
Coding56%
Creative7%
Factual39%
Multilingual38%
Reasoning37%
Rank 67Google Gemini
40%±2
Win rate
Elo 930
Gemini 3 Flash Preview
Tier C
- Won / lost / tied
- 617 / 1245 / 850
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 77.6
By category
Coding34%
Creative21%
Factual50%
Multilingual42%
Reasoning45%
Healthcare23%
Rank 67Z.ai (GLM / Zhipu)
40%±2
Win rate
Elo 928
GLM-5 Turbo
Tier B
- Won / lost / tied
- 378 / 798 / 633
- prompts
- 14
- comparisons
- 1809
- Judge mean
- 54.8
By category
Coding49%
Creative31%
Factual44%
Reasoning24%
Rank 67OpenAI
39%±2
Win rate
Elo 922
gpt-5-2025-08-07
Tier Barchived
- Won / lost / tied
- 432 / 983 / 825
- prompts
- 20
- comparisons
- 2240
- Judge mean
- 57.2
By category
Coding55%
Creative3%
Factual24%
Multilingual37%
Reasoning54%
Rank 67OVH AI Endpoints (GRA)
39%±2
Win rate
Elo 921
gpt-oss-20b
Tier C
- Won / lost / tied
- 518 / 1119 / 748
- prompts
- 20
- comparisons
- 2385
- Judge mean
- 80.1
By category
Coding45%
Creative45%
Factual34%
Multilingual25%
Reasoning54%
Rank 67Z.ai (GLM / Zhipu)
38%±3
Win rate
Elo 916
GLM-5.1
Tier A
- Won / lost / tied
- 372 / 758 / 408
- prompts
- 14
- comparisons
- 1538
- Judge mean
- 69.1
By category
Coding32%
Creative13%
Factual50%
Reasoning43%
Rank 68OVH AI Endpoints (GRA)
37%±2
Win rate
Elo 911
- Won / lost / tied
- 555 / 1254 / 702
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 77.8
By category
Coding49%
Creative10%
Factual29%
Multilingual39%
Reasoning41%
Rank 69Z.ai (GLM / Zhipu)
37%±2
Win rate
Elo 906
GLM-4.5 Air
Tier B
- Won / lost / tied
- 348 / 841 / 559
- prompts
- 14
- comparisons
- 1748
- Judge mean
- 67.4
By category
Coding49%
Creative46%
Factual31%
Reasoning15%
Rank 75Google Gemini
34%±2
Win rate
Elo 885
Gemini Flash Latest
Tier B
- Won / lost / tied
- 559 / 1484 / 669
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 63.7
By category
Coding32%
Creative31%
Factual46%
Multilingual29%
Reasoning31%
Healthcare9%
Rank 75Z.ai (GLM / Zhipu)
33%±2
Win rate
Elo 879
GLM-4.5V (vision)
Tier A
- Won / lost / tied
- 284 / 803 / 451
- prompts
- 14
- comparisons
- 1538
- Judge mean
- 55.9
By category
Coding53%
Creative8%
Factual39%
Reasoning6%
Rank 80OpenAI
28%±2
Win rate
Elo 839
gpt-5-nano-2025-08-07
Tier Barchived
- Won / lost / tied
- 240 / 990 / 516
- prompts
- 18
- comparisons
- 1746
- Judge mean
- 37.9
By category
Coding27%
Creative7%
Factual48%
Multilingual27%
Reasoning6%
Rank 80OVH AI Endpoints (GRA)
28%±2
Win rate
Elo 835
Mistral-7B-Instruct-v0.3
Tier C
- Won / lost / tied
- 441 / 1573 / 497
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 72.5
By category
Coding33%
Creative11%
Factual25%
Multilingual27%
Reasoning30%
Rank 80Z.ai (GLM / Zhipu)
28%±2
Win rate
Elo 835
GLM-4.6
Tier A
- Won / lost / tied
- 221 / 853 / 373
- prompts
- 13
- comparisons
- 1447
- Judge mean
- 52.2
By category
Coding49%
Creative1%
Factual36%
Reasoning5%
Rank 80OVH AI Endpoints (GRA)
26%±2
Win rate
Elo 820
Qwen3.5-397B-A17B
Tier A
- Won / lost / tied
- 294 / 1235 / 464
- prompts
- 19
- comparisons
- 1993
- Judge mean
- 42.5
By category
Coding53%
Creative6%
Factual6%
Multilingual28%
Reasoning4%
Rank 81Google Gemini
24%±2
Win rate
Elo 803
Gemini 3.5 Flash
Tier Aarchived
- Won / lost / tied
- 381 / 1673 / 457
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 51.7
By category
Coding9%
Creative48%
Factual39%
Multilingual25%
Reasoning10%
Rank 82OpenAI
24%±2
Win rate
Elo 798
gpt-5-nano
Tier C
- Won / lost / tied
- 208 / 1154 / 485
- prompts
- 20
- comparisons
- 1847
- Judge mean
- 36.6
By category
Coding18%
Creative4%
Factual50%
Multilingual18%
Reasoning7%
Rank 82Z.ai (GLM / Zhipu)
24%±2
Win rate
Elo 796
GLM-4.6V (vision)
Tier A
- Won / lost / tied
- 179 / 971 / 388
- prompts
- 14
- comparisons
- 1538
- Judge mean
- 47.2
By category
Coding38%
Creative4%
Factual27%
Reasoning5%
Rank 83OVH AI Endpoints (GRA)
23%±2
Win rate
Elo 794
Qwen3-32B
Tier B
- Won / lost / tied
- 426 / 1763 / 322
- prompts
- 20
- comparisons
- 2511
- Judge mean
- 77.5
By category
Coding18%
Creative14%
Factual23%
Multilingual27%
Reasoning27%
Rank 82Google Gemini
23%±2
Win rate
Elo 792
Nano Banana Pro
Tier A
- Won / lost / tied
- 192 / 915 / 292
- prompts
- 18
- comparisons
- 1399
- Judge mean
- 41.4
By category
Coding19%
Creative14%
Factual44%
Multilingual4%
Reasoning3%
Healthcare54%
Rank 84Google Gemini
22%±2
Win rate
Elo 780
- Won / lost / tied
- 250 / 1275 / 353
- prompts
- 24
- comparisons
- 1878
- Judge mean
- 44.1
By category
Coding4%
Creative38%
Factual42%
Multilingual19%
Reasoning13%
Healthcare3%
Rank 84OVH AI Endpoints (GRA)
21%±2
Win rate
Elo 771
Qwen3.5-9B
Tier B
- Won / lost / tied
- 209 / 1338 / 446
- prompts
- 19
- comparisons
- 1993
- Judge mean
- 44.0
By category
Coding36%
Creative5%
Factual11%
Multilingual24%
Reasoning3%
Rank 85Google Gemini
21%±2
Win rate
Elo 768
Gemini 3.1 Pro Preview
Tier C
- Won / lost / tied
- 384 / 1958 / 370
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 48.1
By category
Coding5%
Creative51%
Factual45%
Multilingual16%
Reasoning7%
Healthcare5%
Rank 89Google Gemini
19%±2
Win rate
Elo 750
Gemini Pro Latest
Tier C
- Won / lost / tied
- 338 / 1996 / 378
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 44.4
By category
Coding5%
Creative23%
Factual45%
Multilingual17%
Reasoning8%
Healthcare2%
Rank 92Google Gemini
17%±2
Win rate
Elo 726
Gemini 2.5 Pro
Tier Aarchived
- Won / lost / tied
- 182 / 1437 / 345
- prompts
- 24
- comparisons
- 1964
- Judge mean
- 30.7
By category
Coding17%
Creative3%
Factual34%
Multilingual8%
Reasoning3%
Healthcare54%
Rank 93Google Gemini
15%±1
Win rate
Elo 692
Gemini 2.5 Flash
Tier Aarchived
- Won / lost / tied
- 284 / 2181 / 247
- prompts
- 24
- comparisons
- 2712
- Judge mean
- 37.3
By category
Coding3%
Creative24%
Factual28%
Multilingual11%
Reasoning10%
Healthcare2%
94 models ranked from 111,690 pairwise comparisons across 24 prompts.
Provisional
Too little evidence to rank: needs at least 5 prompts and 30 comparisons.
ProvisionalZ.ai (GLM / Zhipu)
1%±1
Win rate
Elo 155
CogView-4
Tier B
- Won / lost / tied
- 0 / 361 / 6
- prompts
- 4
- comparisons
- 367
- Judge mean
- 0.0
By category
Coding1%
Creative<1% · thin