evals.report
BenchmarksLabsCompareRun guidesIn the wild
BenchmarksReasoning

IMO-Bench

A suite of IMO-level mathematical reasoning benchmarks from Google DeepMind, whose IMO-AnswerBench component tests models on 400 robustified Olympiad problems (Algebra, Combinatorics, Geometry, Number Theory) with verifiable short answers graded by an autograder.

ReasoningaccuracyHigher is better
ScoresAboutRun this benchmark

What is IMO-Bench?

A suite of IMO-level mathematical reasoning benchmarks from Google DeepMind, whose IMO-AnswerBench component tests models on 400 robustified Olympiad problems (Algebra, Combinatorics, Geometry, Number Theory) with verifiable short answers graded by an autograder. evals.report tracks reported IMO-Bench scores with the model, source, status, date, and run caveats attached — official leaderboard scores, vendor-reported launches, and clearly labeled community runs.

Top reported IMO-Bench score: Grok 4 — 73.1% (accuracy).

ModelLabScore↓Source modelStatusDate
Grok 4xAI73.1%—VerifiedJul 9, 2025Details
Gemini 2.5 ProGoogle DeepMind68.2%—VerifiedMar 25, 2025Details
o4-miniOpenAI67.9%—VerifiedApr 16, 2025Details
GPT-5OpenAI65.6%—VerifiedAug 7, 2025Details
o3OpenAI61.1%—VerifiedApr 16, 2025Details
DeepSeek R1OpenDeepSeek60.8%—VerifiedJan 20, 2025Details
Qwen3 235B A22B Instruct 2507OpenAlibaba / Qwen53.8%—VerifiedJul 21, 2025Details
Kimi K2 InstructOpenMoonshot AI45.8%—VerifiedJul 11, 2025Details
DeepSeek V3OpenDeepSeek37.0%—VerifiedDec 26, 2024Details
Claude Sonnet 4Anthropic23.0%—VerifiedMay 22, 2025Details
Claude Opus 4Anthropic22.3%—VerifiedMay 22, 2025Details

Each row reports the model’s accuracy on IMO-Bench. Click a row for the full run context.