HMMT 2025 — Results

Rank	Model	Score
1	granite-4-2-30b	89.17
2	granite-4-2-8b	78.33
3	granite-4-2-3b	66.67
4	gpt-5-2-pro-2025-12-11	1
5	gpt-5-2	0.994
6	deepseek-v3-2-speciale	0.992
7	kimi-k2-thinking-0905	0.975
8	qwen3-6-plus	0.967
9	kimi-k2-5	0.954
10	qwen3-5-397b-a17b	0.948
11	nemotron-3-super-120b-a12b	0.947
12	glm-5-2	0.944
13	glm-5-1	0.94
14	qwen3-6-27b	0.938
15	gpt-5	0.933
16	grok-4-20	0.933
17	qwen3-5-27b	0.92
18	qwen3-5-122b-a10b	0.914
19	qwen3-6-35b-a3b	0.907
20	deepseek-v3-2-thinking	0.902
21	deepseek-v3-2	0.902
22	qwen3-5-35b-a3b	0.89
23	gpt-5-mini	0.878
24	sarvam-105b	0.858
25	mimo-v2-flash	0.844

H

HMMT 2025 — Evaluations

1 phaseActive

No evaluations yet for this benchmark.