Command A

GA

Flagship general model (111B params), 23 languages, tool use/RAG/agents. Context 256k / max output 8k from docs.cohere.com/docs/models.md. No public per-token price: as of 2026-07-26 Cohere publishes no rate for Command A on cohere.com/pricing (no card in the API pricing table, no row in the Model Vault instance table), its docs delegate all figures to that page, and Cohere is absent from the AWS Bedrock Price List API — so this is null rather than the $2.50 in / $10.00 out that aggregators report. Weights are openly released; commercial access appears to be Model Vault / sales-gated.

Command A by Cohere is tracked live in our LLM catalog, with a 256K (256.000-token) context window. It is generally available (GA). Its weights are publicly downloadable, so it can be self-hosted on your own infrastructure.

Last verified: 13 Aug 2026 · sourced from official provider documentation

Provider
Cohere
Status
GA
Input price
Output price
Cached input
Blended price
Context window
256.000 tokens (256K)
Max output
8000 tokens
Modality
text
Open weights
Yes — self-hostable
Knowledge cutoff
Released
Mar 2025
API string
command-a-03-2025

Source: Cohere official documentation ↗

FAQ

Command A — questions & answers

What is the context window of Command A?

Command A has a 256.000-token context window (256K), with up to 8000 output tokens per request.

Is Command A deprecated?

No — Command A is generally available (GA) and not currently scheduled for deprecation or retirement in our tracker.

Does Command A support image or vision input?

Our catalog lists Command A's modalities as text; image/vision input is not among them.

Can I self-host Command A?

Yes — Command A is an open-weight model: its weights are publicly downloadable, so you can run it on your own infrastructure (subject to its license) instead of only calling Cohere's API.

What is the API model string for Command A?

The API model identifier for Command A is "command-a-03-2025" when calling Cohere's API.