Codestral-22B-v0.1-abliterated-v3: benchmark results
Provider: Mistral. Access: Open.
Aggregate Rating 1424 ± 43, rank #4989 of 10916 rated models, from 13 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Chinese LLM - TruthfulQA MC | 56.47 | Accuracy (%) | 70.4 |
| Open Chinese LLM - HellaSwag | 60.95 | Accuracy (%) | 59.9 |
| Open Portuguese LLM - ENEM | 62.77 | Accuracy (%) | 53.4 |
| Open Chinese LLM - ARC Challenge | 52.82 | Accuracy (%) | 51.2 |
| Open Portuguese LLM - OAB Exams | 41.32 | Accuracy (%) | 46.6 |
| Open Portuguese LLM - BLUEX | 49.37 | Accuracy (%) | 44.9 |
| Open Chinese LLM - GSM8K | 43.67 | Accuracy (%) | 44.4 |
| Open Portuguese LLM - ASSIN2 RTE | 88.26 | Macro F1 (%) | 44 |
| Open Portuguese LLM - FaQuAD NLI | 56.61 | Macro F1 (%) | 39.1 |
| Open Chinese LLM - C-Eval Semantic | 64.46 | Accuracy (%) | 37.6 |
| Open Chinese LLM Leaderboard | 55.13 | Official Nonzero-Task Mean (%) | 34.6 |
| Open Chinese LLM - CMMLU | 47.66 | Accuracy (%) | 26.6 |
Interactive version: theaggregate.ai/model?slug=codestral-22b-v0-1-abliterated-v3 · How It Works · Data refreshed daily, snapshot 2026-11-08.