mistral-nemo-minitron-8B-base: benchmark results
NVIDIA's 8B base model width-pruned and distilled from Mistral NeMo 12B, leading its size class on common benchmarks at release. Provider: NVIDIA. Released 2024-08-19. Access: Open.
Aggregate Rating 1433 ± 38, rank #1087 of 1656 rated models, from 52 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval French NLU - FQuAD | 73.34 | Character F1 (%) | 89 |
| EuroEval Spanish NLU - MLQA ES | 66.77 | Character F1 (%) | 88.5 |
| EuroEval Portuguese NLU - MultiWikiQA PT | 75.64 | Character F1 (%) | 87.6 |
| EuroEval Italian NLU - SQuAD IT | 71.28 | Character F1 (%) | 78.9 |
| Open LLM Leaderboard - GPQA | 32.55 | Score | 78.5 |
| EuroEval Portuguese NLU - HAREM | 49.82 | Micro F1 excluding MISC (%) | 76.3 |
| EuroEval Portuguese NLU | 55.3 | NLU Average Score (%) | 75.1 |
| Open LLM Leaderboard - MMLU-Pro | 37.96 | Score | 65.1 |
| EuroEval Portuguese | 51.41 | Average Score (%) | 63.4 |
| EuroEval Danish Knowledge | 68.13 | Knowledge Average Score (%) | 62.3 |
| EuroEval English Knowledge | 79.05 | Knowledge Average Score (%) | 62.3 |
| EuroEval Portuguese NLU - SST-2 PT | 79.07 | Matthews correlation coefficient (%) | 60.9 |
Interactive version: theaggregate.ai/model?slug=mistral-nemo-minitron-8b-base · How It Works · Data refreshed daily, snapshot 2026-10-20.