AIME · 2025 edition · Benchmark definition and scores
AIME · 2025 edition has benchmark disclosures or scores across 11 model releases.
Benchmark versions and metrics
Aliases AIME 2025
- Accuracy · % · Higher is better
Latest models by series · reported scores
The latest models in each named series have no scores to display in this scope. Recorded historical data appears below.
History · 11 releases
- Claude Sonnet 4.61 score record for this release, including 1 exact value (may include historical records)
- Claude Opus 4.61 score record for this release, including 1 exact value (may include historical records)
- GPT-5.22 score records for this release, including 2 exact values (may include historical records)
- Claude Opus 4.52 score records for this release, including 2 exact values (may include historical records)
- Gemini 3 Pro2 score records for this release, including 2 exact values (may include historical records)
- GPT-5.11 score record for this release, including 1 exact value (may include historical records)
- Claude Haiku 4.52 score records for this release, including 2 exact values (may include historical records)
- GPT-51 score record for this release, including 1 exact value (may include historical records)
- GPT-5 mini1 score record for this release, including 1 exact value (may include historical records)
- GPT-5 nano1 score record for this release, including 1 exact value (may include historical records)
- Gemini 2.5 Pro (GA)1 score record for this release, including 1 exact value (may include historical records)