Tiny Aya Global — benchmark results

Cohere Aya-family compact multilingual model. Provider: Cohere. Released 2026-02-17. Access: Open.

Unified ELO 1431 ± 14, rank #1083 of 1776 rated models, from 316 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
EuroEval Albanian Summarization - LR SUM SQ35.09Score (%)99.3
EuroEval Catalan NLU - Guia CAT75.19Sentiment classification Score (%)97.6
EuroEval Italian NLU - SQuAD IT74.62Reading comprehension Score (%)94.4
EuroEval Portuguese NLU - MultiWikiQA PT76.22Reading comprehension Score (%)89.9
EuroEval Spanish Summarization - Mlsum ES28.88Score (%)88.4
EuroEval Slovene NLU - Sentinews52.23Sentiment classification Score (%)87.9
EuroEval Bosnian NLU - MMS BS49.41Sentiment classification Score (%)87.7
EuroEval Bulgarian NLU - MultiWikiQA BG73.92Reading comprehension Score (%)86.3
EuroEval Spanish NLU - MLQA ES65.81Reading comprehension Score (%)86.1
EuroEval Italian Summarization - Ilpost SUM38.14Score (%)84.7
EuroEval Serbian Summarization - LR SUM SR30.08Score (%)83.7
EuroEval Greek Summarization - Greek Wikipedia31.64Score (%)83.6

Interactive version: theaggregate.ai/model?slug=tiny-aya-global · How the rankings work · Data refreshed daily, snapshot 2026-07-22.