Benchgen

BFCL-v3 — Results

RankModelScore
1glm-4-50.778
2glm-4-5-air0.764
3longcat-flash-thinking0.744
4mai-thinking-10.72
5qwen3-next-80b-a3b-thinking0.72
6qwen3-235b-a22b-thinking-25070.719
7qwen3-vl-32b-thinking0.717
8qwen3-235b-a22b-instruct-25070.709
9qwen3-next-80b-a3b-instruct0.703
10qwen3-vl-32b-instruct0.702
11qwen3-vl-30b-a3b-thinking0.686
12qwen3-vl-235b-a22b-instruct0.677
13qwen3-vl-4b-thinking0.673
14qwen3-vl-8b-instruct0.663
15qwen3-vl-30b-a3b-instruct0.663
16qwen3-vl-4b-instruct0.633
17qwen3-vl-8b-thinking0.63
18kimi-k2-instruct0.591
B

BFCL-v3 — Leaderboard

1 phaseActive

BFCL-v3 Leaderboard

No self-reported entries.