AIME 2024

Math% accuracy

American Invitational Mathematics Examination, 2024 problems.

Updated Jun 1, 2025Latest measured Jun 1, 20251 verified · 6 self-reported

Verified results come from third-party or public leaderboard sources. Self-reported results come from provider papers, blogs, or vendor disclosures and should be compared with extra caution.

At a glance

🏆 Top score

o3 OpenAI96.7 % accuracy

Total results

Models tested

Providers

Verified · Self-reported

1 · 6

Average

73.57 % accuracy

Median

79.8 % accuracy

Range

48 – 96.7 % accuracy

Score distribution

Methodology

15 integer-answer problems per exam. Score is percentage of correct answers over the set.

Limitations

Very small problem count per year; high variance. Reasoning-tuned models often dominate.

By provider

OpenAI· 4 models
96.7 % accuracy
o3
Average: 75.82 % accuracyBest: 96.7 % accuracy
xAI· 1 model
83.9 % accuracy
Grok 3
Average: 83.9 % accuracyBest: 83.9 % accuracy
DeepSeek· 1 model
79.8 % accuracy
DeepSeek R1

Full leaderboard

Showing 7 of 7

ProviderSourceSort by

#	Model	Provider	Score (% accuracy)	Source	Date
1	o3	OpenAI	96.7	Self-reported OpenAI o3 system card	Jun 1, 2025
2	Grok 3	xAI	83.9	Self-reported xAI model card	Jun 1, 2025
3	o1	OpenAI	83.3	Self-reported OpenAI o1 system card	Jun 1, 2025
4	DeepSeek R1	DeepSeek	79.8	Self-reported DeepSeek R1 paper	Jun 1, 2025
5	o3-mini	OpenAI	63.3	Self-reported OpenAI system card	Jun 1, 2025
6	GPT-5	OpenAI	60	Self-reported OpenAI system card	Jun 1, 2025
7	Claude Opus 4	Anthropic	48	Third-party Artificial Analysis	Jun 1, 2025

Community ratings

No ratings yet. Be the first to rate AIME 2024.