
Aya 23 8B
Open weights multilingual model covering 23 languages in depth.
About
Aya 23 8B is the smaller of two open weights models released in May 2024 by Cohere for AI, the company's nonprofit research lab. Where the earlier Aya 101 chased breadth across 101 languages, Aya 23 focuses on depth: 23 languages with far more capacity allocated per language, built by pairing a performant pretrained model from the Command family with the large Aya instruction dataset collection.
Cohere reported a 41 percent improvement over Aya 101 on multilingual MMLU and stronger results than Gemma and Mistral class models on discriminative and generative tasks. The 8B variant is aimed at researchers on consumer grade hardware, while a 35B sibling serves heavier workloads. Weights are released under a CC-BY-NC license, so commercial use is restricted, a caveat to check before production plans.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 122; Trending score 80/100 from researcher assessment.