Olmo-3.1-32B-Instruct-SFT: benchmark results
Allen AI's Olmo 3.1 32B Instruct SFT, pretrained on the Dolma 3 dataset and supervised fine-tuned on Dolci data from the Olmo-3-1125-32B base checkpoint. Provider: Allen AI. Released 2025-12-23. Access: Open.
Aggregate Rating 1472 ± 24, rank #879 of 1656 rated models, from 332 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Bosnian NLU - Multi Wiki QA BS | 75.21 | Character F1 (%) | 97.6 |
| EuroEval Croatian NLU - Multi Wiki QA HR | 70.12 | Character F1 (%) | 97.6 |
| EuroEval Romanian NLU - MultiWikiQA RO | 75.08 | Character F1 (%) | 94.7 |
| EuroEval Danish NLU - Multi Wiki QA DA | 80.11 | Character F1 (%) | 93.3 |
| EuroEval English NLU - ScaLA EN | 62.8 | Matthews correlation coefficient (%) | 93.1 |
| EuroEval Catalan NLU - MultiWikiQA CA | 75.28 | Character F1 (%) | 91.9 |
| EuroEval Swedish NLU - Multi Wiki QA SV | 78.34 | Character F1 (%) | 91.7 |
| EuroEval Polish NLU - PoQuAD | 59.18 | Character F1 (%) | 90.6 |
| EuroEval Portuguese NLU | 61.21 | NLU Average Score (%) | 90.5 |
| EuroEval French NLU | 69.82 | NLU Average Score (%) | 89.6 |
| EuroEval German NLU - Germanquad | 64.92 | Character F1 (%) | 89.4 |
| EuroEval Ukrainian NLU - Cross Domain UK Reviews | 61.48 | Matthews correlation coefficient (%) | 88.8 |
Interactive version: theaggregate.ai/model?slug=olmo-3-1-32b-instruct-sft · How It Works · Data refreshed daily, snapshot 2026-10-20.