Benchmark profile

Harvard-MIT Mathematics Tournament 2025 Aggregate (HMMT 2025)

The most recent February edition of the Harvard-MIT Mathematics Tournament, featuring the latest challenging problems in competitive mathematics.

Data verified July 20, 2026

How BenchLM shows HMMT Feb 2025 right now

BenchLM is tracking HMMT Feb 2025 in the local dataset, but exact-source verification records for these rows are still being attached. To avoid a blank benchmark page, BenchLM shows the current tracked rows below as a display-only reference table.

These tracked rows are useful for inspection and spot-checking, but until exact-source attachments are completed they should not be treated as fully verified public benchmark rows.

106 tracked modelsLocal tracked rowsAwaiting exact-source attachmentsDisplay only

Harvard-MIT Mathematics Tournament

Tracked score on HMMT Feb 2025 — July 20, 2026

BenchLM mirrors the published tracked score view for HMMT Feb 2025. GLM-4.7 leads the public snapshot at 97.1% , followed by GPT-5.4 (97%) and GPT-5.2 Pro (97%). BenchLM does not use these results to rank models overall.

GLM-4.7

Z.AI

glm-4-7

97.1%

Overall —

GPT-5.4

OpenAI

gpt-5-4

97%

Overall —

GPT-5.2 Pro

OpenAI

gpt-5-2-pro

97%

Overall —

106 modelsMathCurrentDisplay onlyUpdated July 20, 2026

Tracked score table (106 models)

Score

GLM-4.7Z.AI

97.1%

GPT-5.4OpenAI

97%

GPT-5.2 ProOpenAI

97%

GPT-5.1-Codex-MaxOpenAI

96%

GPT-5.2-CodexOpenAI

96%

GPT-5.3 CodexOpenAI

96%

Grok 4.1xAI

96%

Gemini 3 Pro Deep ThinkGoogle

96%

Claude Opus 4.6Anthropic

96%

GPT-5.1OpenAI

96%

GPT-5.2OpenAI

96%

Claude Sonnet 4.6Anthropic

96%

Gemini 3 ProGoogle

96%

Claude Opus 4.5Anthropic

96%

GPT-5.3 InstantOpenAI

96%

GPT-5.2 InstantOpenAI

96%

Kimi K2.5 (Reasoning)Moonshot AI

95.4%

GLM-5 (Reasoning)Z.AI

95%

GPT-5.3-Codex-SparkOpenAI

95%

Claude Sonnet 4.5Anthropic

94%

Grok 4.1 FastxAI

93%

GPT-5 (high)OpenAI

92%

o1-previewOpenAI

91%

GPT-5 (medium)OpenAI

90%

Qwen3.5 397B (Reasoning)Alibaba

90%

o3-proOpenAI

87%

GPT-5 miniOpenAI

87%

o3OpenAI

85%

GLM-5Z.AI

85%

Grok 4xAI

85%

DeepSeek V3.2 (Thinking)DeepSeek

84%

Qwen2.5-1MAlibaba

82%

Step 3.5 FlashStepFun

82%

Gemini 2.5 ProGoogle

81%

Qwen2.5-72BAlibaba

81%

DeepSeek V3.2DeepSeek

81%

Qwen3.5 397BAlibaba

80%

o4-mini (high)OpenAI

80%

DeepSeek Coder 2.0DeepSeek

78%

Mercury 2Inception

78%

DeepSeekMath V2DeepSeek

77%

DeepSeek LLM 2.0DeepSeek

77%

MiMo-V2-FlashXiaomi

76%

Kimi K2.5Moonshot AI

74%

Claude 4.1 OpusAnthropic

73%

Mistral Large 3Mistral

73%

Aion-2.0Aion Labs

71%

Claude 4 SonnetAnthropic

70%

MiniMax M2.5MiniMax

70%

Seed 1.6ByteDance

69%

Seed-2.0-LiteByteDance

68%

Ministral 3 14B (Reasoning)Mistral

67.5%

Gemini 3 FlashGoogle

67%

Llama 3.1 405BMeta

67%

Claude Haiku 4.5Anthropic

65%

Mistral Large 2Mistral

65%

Ministral 3 14BMistral

65%

Nemotron 3 Super 120B A12BNVIDIA

64%

GPT-4oOpenAI

63%

GLM-4.7-FlashZ.AI

63%

Nemotron 3 Super 100BNVIDIA

62%

Claude 3.5 SonnetAnthropic

62%

Mistral 8x7BMistral

62%

Grok Code Fast 1xAI

61%

Gemini 1.5 ProGoogle

61%

Seed 1.6 FlashByteDance

61%

Gemini 3.1 Flash-LiteGoogle

60%

Gemini 1.0 ProGoogle

59%

Seed-2.0-MiniByteDance

59%

Claude 3 OpusAnthropic

58%

GPT-4 TurboOpenAI

57%

Llama 3 70BMeta

55%

Nemotron 3 Nano 30BNVIDIA

54%

Claude 3 HaikuAnthropic

53%

Nemotron-4 15BNVIDIA

51%

Moonshot v1Moonshot AI

50%

Z-1Z

49%

GPT-OSS 120BOpenAI

48%

Gemini 2.5 FlashGoogle

47%

Nemotron Ultra 253BNVIDIA

46%

Llama 4 BehemothMeta

45%

Llama 4 ScoutMeta

44%

Llama 4 MaverickMeta

43%

LFM2-24B-A2BLiquidAI

43%

Gemma 3 27BGoogle

42%

DeepSeek-R1DeepSeek

41%

Grok 3 [Beta]xAI

39%

Kimi K2Moonshot AI

38.8%

Nova ProAmazon

38%

Qwen3 235B 2507 (Reasoning)Alibaba

37%

Qwen3 235B 2507Alibaba

36%

Claude 4.1 Opus ThinkingAnthropic

35%

GLM-4.5Z.AI

34%

MiniMax M1 80kMiniMax

33%

GLM-4.5-AirZ.AI

32%

DeepSeek V3.1 (Reasoning)DeepSeek

31%

DeepSeek V3.1DeepSeek

30%

Ministral 3 8B (Reasoning)Mistral

30%

GPT-OSS 20BOpenAI

28%

100

Mistral 7B v0.3Mistral

27%

101

Ministral 3 8BMistral

27%

102

Mistral 8x7B v0.2Mistral

26%

103

LFM2.5-1.2B-ThinkingLiquidAI

25%

104

Ministral 3 3B (Reasoning)Mistral

24%

105

LFM2.5-1.2B-InstructLiquidAI

21%

106

Ministral 3 3BMistral

20%

The published HMMT Feb 2025 snapshot places GLM-4.7 first at 97.1%. The third row is 0.1 points behind. The broader top-10 range is 1.1 points, so many of the published results sit in a relatively narrow band.

106 models have been evaluated on HMMT Feb 2025. The benchmark falls in the Math category. This category carries a 5% weight in BenchLM.ai's overall scoring system. HMMT Feb 2025 is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About HMMT Feb 2025

Year

2025

Tasks

Tournament problems

Format

Competition mathematics

Difficulty

High school olympiad level

HMMT Feb 2025 represents the current pinnacle of high school mathematics competition, with problems designed to challenge the brightest mathematical minds.

Harvard-MIT Mathematics Tournament Public benchmark source

BenchLM freshness & provenance

Version

HMMT Feb 2025 2025

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

FAQ

What does HMMT Feb 2025 measure?

The most recent February edition of the Harvard-MIT Mathematics Tournament, featuring the latest challenging problems in competitive mathematics.

Which model leads the published HMMT Feb 2025 snapshot?

GLM-4.7 currently leads the published HMMT Feb 2025 snapshot with 97.1% tracked score. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on HMMT Feb 2025?

106 AI models are included in BenchLM's mirrored HMMT Feb 2025 snapshot, based on the public leaderboard captured on July 20, 2026.

Learn More

Read our explainer: HMMT Feb 2025 benchmark deep dive

Last updated: July 20, 2026 · mirrored from the public benchmark leaderboard

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.