h2o-danube2-1.8B-sft: benchmark results
Provider: H2O.ai. Access: Open.
Unified ELO 1318 ± 22, rank #2629 of 2928 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 25.63 | Accuracy (%) (5-shot) | 44.1 |
| Open Portuguese LLM - FaQuAD NLI | 54.55 | Macro F1 (%) | 36.8 |
| Open Portuguese LLM - ASSIN2 RTE | 70.62 | Macro F1 (%) | 25.2 |
| Open Portuguese LLM - OAB Exams | 28.34 | Accuracy (%) | 23.7 |
| Open LLM Leaderboard v1 - HellaSwag | 72.75 | Normalized accuracy (%) (10-shot) | 23.2 |
| Open LLM Leaderboard v1 - WinoGrande | 68.51 | Accuracy (%) (5-shot) | 22.7 |
| Open Portuguese LLM - ENEM | 25.26 | Accuracy (%) | 21.2 |
| Open Portuguese LLM - BLUEX | 24.9 | Accuracy (%) | 19.6 |
| Open LLM Leaderboard v1 - ARC Challenge | 42.66 | Normalized accuracy (%) (25-shot) | 18.7 |
| Open LLM Leaderboard v1 - MMLU | 35.9 | Accuracy (%) (5-shot) | 18.6 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 38.7 | MC2 (%) (0-shot) | 8.7 |
Interactive version: theaggregate.ai/model?slug=h2o-danube2-1-8b-sft · How It Works · Data refreshed daily, snapshot 2026-09-23.