Benchgen

SuperGPQA — Results

RankModelScore
1deepseek-v4-1-flash53.1
2aliceai-foundation-80b-a3b-base44.3
3minicpm5-2b40.8
4qwen3-7-max0.736
5qwen3-6-plus0.716
6qwen3-7-plus0.714
7seed-2-1-pro0.708
8qwen3-5-397b-a17b0.704
9seed-2-1-turbo0.674
10qwen3-5-122b-a10b0.671
11qwen3-6-27b0.66
12qwen3-5-27b0.656
13qwen3-235b-a22b-thinking-25070.649
14qwen3-6-35b-a3b0.647
15qwen3-vl-235b-a22b-thinking0.643
16qwen3-5-35b-a3b0.634
17qwen3-235b-a22b-instruct-25070.626
18qwen3-next-80b-a3b-thinking0.608
19qwen3-vl-235b-a22b-instruct0.604
20qwen3-vl-32b-thinking0.59
21qwen3-next-80b-a3b-instruct0.588
22qwen3-5-9b0.582
23kimi-k2-instruct-09050.572
24kimi-k2-instruct0.572
25qwen3-vl-30b-a3b-thinking0.564
S

SuperGPQA — Leaderboard

1 phaseActive

SuperGPQA Leaderboard

No self-reported entries.