Best Expert: benchmark results
Reference assembled from reported best expert and contest performances across different participants; it does not describe one person's measured skill profile. Provider: Human.
Aggregate Rating 2015 ± 30, from 24 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA AIME 2025 | 100 | Accuracy (%) | 100 |
| ALE-Bench | 2341 | Performance (Self-Refine x1) (imported) | 100 |
| ARC-AGI-2 | 100 | Accuracy (%) | 100 |
| ArchSIBench | 92.3 | Avg. (self-reported) | 100 |
| Chess Bench LLM | 2236 | Lichess-Classical-calibrated Glicko-2 rating (native signed | 100 |
| CodeElo | 2157 | Elo Rating | 100 |
| LLM Stats (AIME 2024) | 100 | Score (%) | 100 |
| MATH (CoT) | 90 | Score (%) | 100 |
| MMMU Benchmark | 88.6 | Validation Score | 100 |
| MathArena - IMC 2025 | 100 | Accuracy (%) | 100 |
| MathArena - IMO 2025 | 100 | Accuracy (%) | 100 |
| MathArena - USAMO 2025 | 100 | Accuracy (%) | 100 |
Interactive version: theaggregate.ai/model?slug=best-expert · How It Works · Data refreshed daily, snapshot 2026-10-20.