Aya Expanse

Name: Aya Expanse
Author: Cohere

CohereOpen WeightsPending Human Review

Aya Expanse is a multilingual text generation family with 8B and 32B variants, targeting strong instruction following across 23 languages. Its training combines data arbitrage, multilingual preference optimization, and model merging. The 32B model supports a 128K context, while the smaller 8B release supports 8K. Research weights are downloadable under non-commercial terms. Cohere’s current Chat API serves the 32B checkpoint with a 4K output limit; the hosted 8B checkpoint was retired in April 2026. This family page groups the two sizes without treating their context limits or current API access as identical.

2024-10-24

8B and 32B

Dense autoregressive Transformer

CC-BY-NC-4.0 with acceptable-use addendum

Compare with other models

Specifications

Parameters: 8B and 32B
Architecture: Dense autoregressive Transformer
License: CC-BY-NC-4.0 with acceptable-use addendum
Context Window: 128,000 tokens
Max Output: 4,000 tokens
Type: text
Modalities: text

Benchmark Scores

Advanced Specifications

Model Family: Aya
API Access: Available
Chat Interface: Available
Multilingual Support: Yes
Variants: aya-expanse-8b: 8K context; weights availableaya-expanse-32b: 128K context; current hosted API

Capabilities & Limitations

Capabilities: 23-language supportinstruction followingtranslationsummarizationlong context
Known Limitations: Downloaded weights are licensed for non-commercial use; commercial deployments require separate terms.The 8B variant has an 8K context and its hosted API was retired April 4, 2026.128K context and current API availability refer to the 32B variant.
Notable Use Cases: multilingual assistantsglobal customer supportmultilingual preference research

Resources

Related Models

North Small Translate

Cohere

North Small Translate is a dedicated text translation model supporting more than 50 languages. Its 218B-parameter mixture-of-experts network activates 25B parameters and targets long-form translation with efficient serving. Open weights are available in several precisions under CC BY-NC 4.0, while commercial use requires appropriate licensing. It can be evaluated through Cohere’s Chat API and a public demo. The release date follows the September 9 API changelog; the launch blog followed on September 10. Translation quality claims depend on the specific language pair, domain, and evaluation method.

Typetext

Parameters218B total, 25B active

2026-09-09

Open Weights

Details Compare

Tiny Aya Thinker

Cohere

Tiny Aya Thinker is Cohere Labs’ 3.35B multilingual reasoning research release, published with the Building Multilingual Bridges paper. L2-Thinker produces reasoning in the user’s language across 60 languages, while En-Thinker provides an English-reasoning comparison. The models build on a 32K Tiny Aya base checkpoint. Downloadable research weights and a public demonstration support multilingual reasoning experiments. The paper publication precedes the October 5 explanatory blog; that later blog is not the first release.

North Micro Vision Instruct

Cohere

North Micro Vision Instruct is a compact open-weight vision-language model for captioning, visual questions, grounding, and document understanding. It combines a 2B language model with a 400M vision encoder and preserves image aspect ratios through native-resolution processing. The language backbone supports 128K tokens, but multimodal training and validated operating conditions cover only 8K; this page uses the conservative multimodal limit. Apache-licensed weights make it a foundation for domain-specific fine-tuning and local experimentation. It supports interleaved images and text, multilingual prompts, and text output.