Atlas

Benchmarks

← All benchmarks

HMMT Nov 2025

Math · 2025-11-17

This MathArena track evaluates models on the HMMT Nov 2025 problem set. Scores report the percentage of problems answered correctly.

Top models (higher is better)

ModelScore
GPT-5.299.2
GLM-594.2
Step 3.5 Flash94.2
DeepSeek-V3.2-Speciale93.3
Gemini 3 Flash Preview93.3
Gemini 3 Pro Preview93.3
Grok 4.1 Fast93.3
GLM 4.691.7
GPT-5.191.7
Grok 4 Fast90.8
DeepSeek-V3.290.0
gpt-oss-120b90.0
GPT-589.2
Kimi K2.589.2
Kimi K2 Thinking89.2
Grok 488.3
DeepSeek-V3.2-Exp84.2
GPT-5 Mini84.2
GPT-5 Nano81.7
Gemini 2.5 Pro80.0
Falcon H1R 7B79.2
QED-Nano75.0
Loading Atlas data…