canarim-7B: benchmark results
Provider: Other. Access: Open.
Unified ELO 1337 ± 21, rank #2556 of 2928 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Portuguese LLM - FaQuAD NLI | 49.09 | Macro F1 (%) | 31.4 |
| Open LLM Leaderboard v1 - HellaSwag | 77.52 | Normalized accuracy (%) (10-shot) | 31.1 |
| Open LLM Leaderboard v1 - GSM8K | 9.93 | Accuracy (%) (5-shot) | 30.9 |
| Open Portuguese LLM - OAB Exams | 31.48 | Accuracy (%) | 28.3 |
| Open LLM Leaderboard v1 - WinoGrande | 71.43 | Accuracy (%) (5-shot) | 26.6 |
| Open LLM Leaderboard v1 - ARC Challenge | 51.96 | Normalized accuracy (%) (25-shot) | 26.5 |
| Open Portuguese LLM - ASSIN2 RTE | 71.96 | Macro F1 (%) | 26 |
| Open Portuguese LLM - BLUEX | 29.76 | Accuracy (%) | 25.6 |
| Open Portuguese LLM - ENEM | 25.96 | Accuracy (%) | 21.5 |
| Open LLM Leaderboard v1 - MMLU | 40.92 | Accuracy (%) (5-shot) | 21.4 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 40.03 | MC2 (%) (0-shot) | 12.8 |
Interactive version: theaggregate.ai/model?slug=canarim-7b · How It Works · Data refreshed daily, snapshot 2026-09-23.