Command A
Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding use cases.
command-a-03-2025Released 13 Mar 2025 (Mar 2025)Pricing
List prices at Cohere Chat, in US dollars per million tokens.
| Tier | Input / 1M | Output / 1M | |
|---|---|---|---|
| Standard | $2.50 | $10.00 |
What it costs a month
At Cohere Chat list prices, 30 days.
- Chat assistant5,000 requests a day, 2K in / 0.5K out$1,500
- RAG over documents2,000 requests a day, 8K in / 0.6K out$1,560
- Coding agent1,000 requests a day, 40K in / 2K out$3,600
- Overnight batch20,000 requests a day, 4K in / 0.4K out$8,400
Where to run it
3 hosts serve Command A. The comparison is on blended price, 3 parts input to 1 part output.
| Host | In / out per 1M | Model id | |
|---|---|---|---|
| $2.50 / $10.00 | command-a-03-2025 | ||
| $1.56 / $1.56 −64% vs headline | cohere.command-a-03-2025 | ||
Vercel AI GatewayGateway command-a | $2.50 / $10.00 | command-a |
Capabilities
Core
- Streaming
- Function calling
- JSON mode
- Structured outputs
Multimodal
- Vision
- Image input
- Audio
- Image generation
- Video
Advanced
- Reasoning
- Tool calling
- MCP
- Prompt caching
- Batch API
- Embeddings
- Takes in
- Text
- Produces
- Text
- Open weights
- License: CC-BY-NC-4.0Runs in about 266.5 GB of GPU memory
Cheaper alternatives
Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.
Same model, cheaper host
- Oracle OCI64% lesscohere.command-a-03-2025$1.56 / $1.56
Smaller Cohere models
No cheaper Cohere model keeps its key capabilities.
Other frontier labs
- 9% less
- 9% less
- 9% less
- 31% less
Open-weight picks
- 9% less
- 31% less
- 61% less
- 80% less
Call it
Using Cohere Chat's API.
from litellm import completion
response = completion(
model="cohere_chat/command-a-03-2025",
messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)Quick answers
How much does Command A cost?
On Cohere Chat, Command A costs $2.50 per million input tokens and $10.00 per million output tokens. The cheapest host is Oracle OCI (2025-03), at $1.56 per million input tokens and $1.56 per million output tokens.
Which APIs offer Command A?
Command A runs on 3 APIs: Cohere Chat, Oracle OCI and Vercel AI Gateway.
What is Command A's context window?
256K tokens, with up to 8K tokens in one reply.
Is Command A being retired?
No retirement date is scheduled for it on any API we track.
What are cheaper alternatives to Command A?
Models that cost less and keep most of what it can do: GPT-6.2 Sol on OpenAI (9% less), Gemini 4 Carbon on Google AI Studio (9% less) and Kimi K3.1 Pro on Databricks (9% less). Command A itself costs 64% less on Oracle OCI (2025-03). Savings are on a mix of 3 input tokens to 1 output token.
When was Command A released?
Cohere released it on 13 Mar 2025.
What is Command A's API model id?
Use command-a-03-2025 with Cohere Chat. Other hosts use their own ids: cohere.command-a-03-2025 on Oracle OCI, command-a on Vercel AI Gateway.
Sources
Raw catalog entry for command-a-03-2025 on Cohere Chat
{
"input_cost_per_token": 0.0000025,
"litellm_provider": "cohere_chat",
"max_input_tokens": 256000,
"max_output_tokens": 8000,
"mode": "chat",
"output_cost_per_token": 0.00001,
"supports_function_calling": true,
"supports_response_schema": true,
"supports_tool_choice": true
}