mpt-7B-8k: benchmark results
Provider: Databricks. Access: Open.
Unified ELO 1338 ± 20, rank #2546 of 2928 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Portuguese LLM - ASSIN2 RTE | 81.12 | Macro F1 (%) | 31.5 |
| Open LLM Leaderboard v1 - HellaSwag | 77.4 | Normalized accuracy (%) (10-shot) | 30.8 |
| Open LLM Leaderboard v1 - GSM8K | 8.34 | Accuracy (%) (5-shot) | 28.4 |
| Open LLM Leaderboard v1 - WinoGrande | 71.11 | Accuracy (%) (5-shot) | 26 |
| Open Portuguese LLM - ENEM | 32.05 | Accuracy (%) | 25.4 |
| Open Portuguese LLM - BLUEX | 28.09 | Accuracy (%) | 23.9 |
| Open LLM Leaderboard v1 - ARC Challenge | 47.35 | Normalized accuracy (%) (25-shot) | 22.3 |
| Open LLM Leaderboard v1 - MMLU | 42.58 | Accuracy (%) (5-shot) | 22.1 |
| Open Portuguese LLM - OAB Exams | 27.02 | Accuracy (%) | 21.2 |
| Open Portuguese LLM - FaQuAD NLI | 43.97 | Macro F1 (%) | 17.1 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 36.65 | MC2 (%) (0-shot) | 4.1 |
Interactive version: theaggregate.ai/model?slug=mpt-7b-8k · How It Works · Data refreshed daily, snapshot 2026-09-23.