Grok 3 Mini (High): benchmark results
Grok 3 Mini evaluated at the high reasoning-effort setting. Provider: xAI. Released 2025-02-17. Access: API.
Aggregate Rating 1645 ± 17, rank #539 of 2168 rated models, from 71 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MATH-Perturb (Hard) | 87.1 | Accuracy (%) | 94.4 |
| BenchTable - STEM | 80.1 | Weighted Score (%) | 92.2 |
| BenchTable - Utility | 72.6 | Weighted Score (%) | 88.4 |
| Olfactory Perception Benchmark - Rate-All-That-Apply | 37 | Score (%) on rate-all-that-apply descriptor profiling (selec | 85 |
| Confabulation Leaderboard (Lechmazur) | 6.93 | Confabulation rate % (lower is better) | 83.1 |
| Vals AI MATH 500 | 94.2 | Accuracy (%) | 82.9 |
| BenchTable | 67.3 | Total Score (%) | 82.8 |
| Olfactory Perception Benchmark - Odor Pleasantness | 73.7 | Score (%) on odor pleasantness paired comparison (which of t | 80 |
| Step Game (Lechmazur) | 3.24 | TrueSkill μ | 78.4 |
| MATH Level 5 | 88.07 | Accuracy (%) | 76.6 |
| BenchTable - Tech | 70.7 | Weighted Score (%) | 71.5 |
| PlatinumBench (MIT) | 1.51 | Category-balanced mean error rate (%) | 69.7 |
Interactive version: theaggregate.ai/model?slug=grok-3-mini-high · How It Works · Data refreshed daily, snapshot 2026-10-20.