command-a-03-2025

cc-by-nc-4.0

Command A is our most performant model to date, excelling at tool use, agents, retrieval augmented generation (RAG), and multilingual use cases. Command A has a context length of 256K, only requires two GPUs to run, and has 150% higher throughput compared to Command R+ 08-2024.

Context
256K
Max output
8K
Input price
$1/1M tokens
Output price
$2/1M tokens

Capabilities

Reasoning & Interaction
Tool Calling
Infrastructure
Streaming
Fine-tuning

Details

Provider Cohere
Creator cohere
Familycommand-a
Licensecc-by-nc-4.0
Statusactive
Typechat
Input modalitiestext
Output modalitiestext
Open weightYes
Parameters—
Architecture—
Context window256K tokens
Max input—
Max output8K tokens
Reasoning tokens—
Knowledge cutoff—
Release date—
Sourceofficial

Tools

Function Callingfunction_calling
Call external functions and APIs

Endpoints

chatPOST
https://api.cohere.com/v2/chat

Pricing

Input
$1
Output
$2
Cache write
—
Cache read
—
Batch in
—
Batch out
—

Family Comparison: command-a

ModelContextMax outPricing
command-a-03-2025256K8K$1$2
command-a-plus-05-2026128K64K$1$2
command-a-reasoning-08-2025256K32K$1$2
command-a-translate-08-20258K8K$1$2
command-a-vision-07-2025128K8K$1$2

API

GET/v1/models/cohere/command-a-03-2025