HMMT 2025 — Results

Rank	Model	Score
1	granite-4-2-30b	89.17
2	twil-8b	86.7
3	granite-4-2-8b	78.33
4	granite-4-2-3b	66.67
5	gpt-5-2-pro-2025-12-11	1
6	gpt-5-2	0.994
7	deepseek-v3-2-speciale	0.992
8	kimi-k2-thinking-0905	0.975
9	qwen3-6-plus	0.967
10	kimi-k2-5	0.954
11	qwen3-5-397b-a17b	0.948
12	nemotron-3-super-120b-a12b	0.947
13	glm-5-2	0.944
14	glm-5-1	0.94
15	qwen3-6-27b	0.938
16	gpt-5	0.933
17	grok-4-20	0.933
18	qwen3-5-27b	0.92
19	qwen3-5-122b-a10b	0.914
20	qwen3-6-35b-a3b	0.907
21	deepseek-v3-2-thinking	0.902
22	deepseek-v3-2	0.902
23	qwen3-5-35b-a3b	0.89
24	gpt-5-mini	0.878
25	sarvam-105b	0.858

H

HMMT 2025 — Evaluations

1 phaseActive

No evaluations yet for this benchmark.