Trinity-2-Codestral-22B: benchmark results
Migel Tissera's Trinity-2 coding model, a fine-tune of Mistral's Codestral 22B with compute sponsored by KindoAI. Provider: Arcee AI. Released 2024-08-07. Access: Open.
Unified ELO 1519 ± 1, rank #662 of 1553 rated models, from 7 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM Leaderboard | 46.28 | Average Score (%) | 92.9 |
| Open LLM Leaderboard - BBH | 36.41 | Score | 74.9 |
| Open LLM Leaderboard - GPQA | 8.61 | Score | 71.4 |
| Open LLM Leaderboard - MuSR | 9.61 | Score | 46.9 |
| Open LLM Leaderboard - MMLU-Pro | 25.64 | Score | 46.4 |
| Open LLM Leaderboard - MATH Level 5 | 9.67 | Score | 46.2 |
| Open LLM Leaderboard - IFEval | 42.02 | Score | 43 |
Interactive version: theaggregate.ai/model?slug=trinity-2-codestral-22b · How It Works · Data refreshed daily, snapshot 2026-09-08.