aya-expanse-32B: benchmark results
Cohere Aya Expanse 32B multilingual model. Provider: Cohere. Released 2024-10-24. Access: Open.
Aggregate Rating 1527 ± 9, rank #692 of 1656 rated models, from 571 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| CommunityFact (Closed-Input) - Finance (English) | 69.7 | Macro-F1 (%) over True and False claims on the temporally he | 100 |
| CommunityFact (Closed-Input) - Politics (English) | 71.79 | Macro-F1 (%) over True and False claims on the temporally he | 100 |
| CommunityFact (Closed-Input) - Politics (Spanish) | 69.37 | Macro-F1 (%) over True and False claims on the temporally he | 100 |
| Open Arabic LLM - Alghafa Multiple Choice Facts Truefalse Balanced Task | 98.67 | Accuracy (%) | 100 |
| Open Japanese LLM - ALT E TO J Comet Wmt22 | 91.67 | Score (%) | 98.6 |
| Open CoT - LSAT Reading Comprehension | 24.54 | CoT Gain (%) | 98.5 |
| MT-Bench PL - Roleplay | 9.7 | Judge Score (0-10) | 98 |
| MT-Bench PL - Writing | 9.6 | Judge Score (0-10) | 98 |
| Open Japanese LLM - ALT J TO E Bleu EN | 19.43 | Score (%) | 97.7 |
| Open Japanese LLM - ALT E TO J Bert Score JA F1 | 87.55 | Score (%) | 97.5 |
| Open PL LLM - PPC (generative, 5-shot) | 80.8 | Accuracy (%) | 97 |
| Open Japanese LLM - SUM | 13.88 | Score (%) | 96.4 |
Interactive version: theaggregate.ai/model?slug=aya-expanse-32b · How It Works · Data refreshed daily, snapshot 2026-10-20.