Codestral 22B: benchmark results
Provider: Mistral.
Aggregate Rating 1482 ± 30, rank #2638 of 10916 rated models, from 7 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LilyBench - Zero-Shot Compile Rate | 79.3 | Compile rate (%) of zero-shot LilyPond generations over a 20 | 100 |
| Beyond Pass Rate | 12.3 | Correctness (%) | 50 |
| SolBehave - Compile@10 - One Repair | 39.9 | Compiling samples (%) | 16.7 |
| SolBehave - Compile@10 - Raw | 34.3 | Compiling samples (%) | 16.7 |
| LLM Stats Score | -0.44 | LLM Stats Score (conservative rating) | 10.7 |
| LiveSQLBench | 12.53 | Success Rate (imported) | 9.4 |
| LilyBench - Understanding | 47.4 | Macro average (0-1 scaled to %) over the eight Mutopia under | 0 |
Interactive version: theaggregate.ai/model?slug=codestral-22b · How It Works · Data refreshed daily, snapshot 2026-11-08.