aya-expanse-8B: benchmark results
Cohere Aya Expanse 8B multilingual model. Provider: Cohere. Released 2024-10-24. Access: Open.
Aggregate Rating 1460 ± 10, rank #931 of 1656 rated models, from 553 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LA Leaderboard - PAWS-X Spanish | 66.95 | Accuracy (%) | 100 |
| LA Leaderboard - ParafrasesGL | 58.5 | Accuracy (%) | 100 |
| LA Leaderboard - XL-Sum Spanish | 4.33 | BLEU | 100 |
| LA Leaderboard - TELEIA | 66.67 | Accuracy (%) | 98.5 |
| LatamBoard - FLORES Bidirectional | 46.97 | Score (%) | 97 |
| LatamBoard - Translation Score | 48.24 | Score (%) | 97 |
| LA Leaderboard - XQuAD Spanish | 77.98 | F1 (%) | 96.9 |
| LA Leaderboard - WNLI Basque | 60.56 | Accuracy (%) | 95.6 |
| LA Leaderboard - XNLI Galician | 49.77 | Accuracy (%) | 95.6 |
| LA Leaderboard - caBREU Catalan | 18.55 | BLEU | 95.6 |
| LA Leaderboard - Summarization Galician | 7.26 | BLEU | 94 |
| CommunityFact (Closed-Input) - Finance (French) | 65.52 | Macro-F1 (%) over True and False claims on the temporally he | 92.9 |
Interactive version: theaggregate.ai/model?slug=aya-expanse-8b · How It Works · Data refreshed daily, snapshot 2026-10-20.