Command A Reasoning

Name: Command A Reasoning
Author: Cohere

CohereOpen WeightsPending Human Review

Command A Reasoning is Cohere’s first dedicated reasoning language model, extending the enterprise Command A generation with controllable thinking. Its 111B-parameter dense Transformer supports 23 languages, tool-driven workflows, and document-grounded responses. Applications can switch thinking on or off and set a reasoning budget to trade computation for task performance. The current hosted model provides a 256K context and up to 32K output tokens. Public research weights use non-commercial terms, with commercial inference available separately. Its reasoning specialization is a distinct catalog identity from the standard Command A model and later sparse Command A+ release.

2025-08-21

111B

Dense decoder-only Transformer with local and global attention

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

Compare with other models

Specifications

Parameters: 111B
Architecture: Dense decoder-only Transformer with local and global attention
License: CC-BY-NC-4.0 with acceptable-use addendum
Context Window: 256,000 tokens
Max Output: 32,000 tokens
Training Data Cutoff: June 1, 2024
Type: text
Modalities: text

Benchmark Scores

Advanced Specifications

Model Family: Command
API Access: Available
Chat Interface: Available
Multilingual Support: Yes
Variants: command-a-reasoning-08-2025

Capabilities & Limitations

Capabilities: controllable reasoningtool useretrieval augmented generationmultilingualcode
Known Limitations: Downloaded weights are licensed for non-commercial use; commercial deployments require separate terms.Thinking increases latency and token consumption.Reasoning traces and citations do not guarantee correct conclusions.
Notable Use Cases: enterprise reasoning agentscomplex document questionsmulti-step workflow automation
Function Calling Support: Yes
Tool Use Support: Yes

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.