h2o-danube-1.8B-base: benchmark results
Provider: H2O.ai. Access: Open.
Unified ELO 1232 ± 21, rank #2839 of 2928 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - HellaSwag | 69.58 | Normalized accuracy (%) (10-shot) | 20.2 |
| Open LLM Leaderboard v1 - WinoGrande | 64.48 | Accuracy (%) (5-shot) | 17.4 |
| Open Portuguese LLM - ASSIN2 RTE | 53.78 | Macro F1 (%) | 17.3 |
| Open Portuguese LLM - FaQuAD NLI | 43.97 | Macro F1 (%) | 17.1 |
| Open LLM Leaderboard v1 - GSM8K | 1.44 | Accuracy (%) (5-shot) | 17 |
| Open Portuguese LLM - ENEM | 21.06 | Accuracy (%) | 16.3 |
| Open LLM Leaderboard v1 - ARC Challenge | 39.42 | Normalized accuracy (%) (25-shot) | 15.8 |
| Open Portuguese LLM - BLUEX | 22.67 | Accuracy (%) | 15 |
| Open Portuguese LLM - OAB Exams | 24.01 | Accuracy (%) | 11.7 |
| Open LLM Leaderboard v1 - MMLU | 25.94 | Accuracy (%) (5-shot) | 10.6 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 33.86 | MC2 (%) (0-shot) | 0.8 |
Interactive version: theaggregate.ai/model?slug=h2o-danube-1-8b-base · How It Works · Data refreshed daily, snapshot 2026-09-23.