Human Preference Ranking on Arena Human Preference 140K
1Median Rankgemini-2.5-pro
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| gemini-2.5-pro2026.06 | 1 | 1 | 1 | 1,126.47 | 1,120.09 | |
| o3-2025-04-16Release Date=2025-04-162026.06 | 2 | 2 | 2 | 1,082.69 | 1,076.17 | |
| chatgpt-4o-latest-20250326Release Date=202503262026.06 | 3 | 2 | 2 | 1,081.35 | 1,074.47 | |
| gemini-2.5-pro-preview-05-06Mode=preview, Release Date=05-062026.06 | 4 | 2 | 2 | 1,077.86 | 1,067.67 | |
| deepseek-r1-0528Release Date=05282026.06 | 5 | 2 | 2 | 1,076.53 | 1,069.29 | |
| grok-3-preview-02-24Mode=preview, Release Date=02-242026.06 | 6 | 2 | 2 | 1,074.11 | 1,066.47 | |
| llama-4-maverick-03-26-experimentalMode=experimental, Release Date=03-262026.06 | 7 | 2 | 2 | 1,069.55 | 1,061.76 | |
| gemini-2.5-flash2026.06 | 8 | 4 | 4 | 1,063.81 | 1,057.94 | |
| qwen3-235b-a22b-no-thinkingThinking=false, Parameter Count=235B2026.06 | 9 | 9 | 9 | 1,044.47 | 1,038.45 | |
| gemini-2.5-flash-preview-04-17Mode=preview, Release Date=04-172026.06 | 10 | 9 | 9 | 1,037.17 | 1,028.98 | |
| kimi-k2-0711-previewMode=preview, Release Date=07112026.06 | 11 | 9 | 9 | 1,033.42 | 1,022.11 | |
| gpt-4.1-2025-04-14Release Date=2025-04-142026.06 | 12 | 10 | 9 | 1,030.27 | 1,023.1 | |
| deepseek-v3-0324Release Date=03242026.06 | 13 | 10 | 9 | 1,029.53 | 1,022.22 | |
| qwen-max-2025-01-25Release Date=2025-01-252026.06 | 14 | 10 | 9 | 1,028.6 | 1,019.99 | |
| qwen3-235b-a22bParameter Count=235B2026.06 | 15 | 10 | 10 | 1,027.72 | 1,019.79 | |
| mistral-medium-2505Release Date=25052026.06 | 16 | 10 | 10 | 1,025.82 | 1,019.87 | |
| o4-mini-2025-04-16Release Date=2025-04-162026.06 | 17 | 10 | 10 | 1,025.68 | 1,018.49 | |
| gemini-2.5-flash-lite-preview-06-17-thinkingMode=preview, Release Date=06-17, Thinking=true2026.06 | 18 | 10 | 10 | 1,024.24 | 1,016.27 | |
| minimax-m12026.06 | 19 | 11 | 11 | 1,016.65 | 1,008.83 | |
| gemma-3-27b-itParameter Count=27B, Type=Instruct2026.06 | 20 | 14 | 11 | 1,015.18 | 1,008.43 | |
| claude-opus-4-20250514-thinking-16kRelease Date=20250514, Thinking Budget=16k2026.06 | 21 | 18 | 17 | 1,009.65 | 1,002.18 | |
| grok-3-mini-betaMode=mini-beta2026.06 | 22 | 18 | 17 | 1,009.41 | 1,001.98 | |
| claude-opus-4-20250514Release Date=202505142026.06 | 23 | 19 | 19 | 1,007.92 | 1,002.18 | |
| grok-3-mini-highMode=mini-high2026.06 | 24 | 19 | 17 | 1,005.02 | 994.19 | |
| claude-sonnet-4-20250514-thinking-32kRelease Date=20250514, Thinking Budget=32k2026.06 | 25 | 24 | 24 | 992.34 | 984.56 | |
| gemini-2.0-flash-001Version=0012026.06 | 26 | 24 | 24 | 991.42 | 983.97 | |
| qwq-32bParameter Count=32B2026.06 | 27 | 24 | 24 | 989.6 | 980.74 | |
| mistral-small-2506Release Date=25062026.06 | 28 | 24 | 21 | 987.95 | 976.29 | |
| gpt-4.1-mini-2025-04-14Release Date=2025-04-142026.06 | 29 | 24 | 24 | 987.03 | 979.73 | |
| qwen3-30b-a3bParameter Count=30B2026.06 | 30 | 25 | 25 | 981.07 | 973.4 | |
| command-a-03-2025Release Date=03-20252026.06 | 31 | 25 | 25 | 977.89 | 971.06 | |
| claude-sonnet-4-20250514Release Date=202505142026.06 | 32 | 27 | 25 | 977.04 | 970.66 | |
| o3-mini2026.06 | 33 | 33 | 33 | 961.53 | 954.33 | |
| amazon-nova-experimental-chat-05-14Mode=experimental, Release Date=05-142026.06 | 34 | 33 | 32 | 959.25 | 949.2 | |
| gemma-3n-e4b-itType=Instruct2026.06 | 35 | 33 | 32 | 957.85 | 946.51 | |
| llama-4-scout-17b-16e-instructParameter Count=17B, Type=Instruct2026.06 | 36 | 34 | 34 | 939.81 | 928.81 | |
| claude-3-7-sonnet-20250219-thinking-32kRelease Date=20250219, Thinking Budget=32k2026.06 | 37 | 35 | 35 | 939.67 | 932.63 | |
| llama-4-maverick-17b-128e-instructParameter Count=17B, Type=Instruct2026.06 | 38 | 35 | 35 | 939.06 | 930.94 | |
| claude-3-7-sonnet-20250219Release Date=202502192026.06 | 39 | 36 | 36 | 933.81 | 926.53 | |
| claude-3-5-sonnet-20241022Release Date=202410222026.06 | 40 | 36 | 36 | 932.98 | 925.73 | |
| llama-3.3-70b-instructParameter Count=70B, Type=Instruct2026.06 | 41 | 41 | 41 | 914.06 | 905.44 | |
| amazon.nova-pro-v1:0Version=v1:02026.06 | 42 | 41 | 41 | 913.6 | 906.28 | |
| mistral-small-3.1-24b-instruct-2503Parameter Count=24B, Type=Instruct, Release Date=25032026.06 | 43 | 41 | 41 | 910.09 | 899.89 | |
| claude-3-5-haiku-20241022Release Date=202410222026.06 | 44 | 41 | 41 | 900.34 | 892.9 | |
| magistral-medium-2506Release Date=25062026.06 | 45 | 43 | 44 | 890.5 | 878.45 |