Very long inputs
Choose Command A
Command A reads up to 256k tokens per request, against 128k for Codestral.
Codestral vs Command A: price, context and capabilities
Choose Codestral for the latest release. Choose Command A for very long inputs.
In short
- A price comparison is not possible: one of them has no comparable published price.
- Command A reads more per request: 256k tokens against 128k.
- Codestral is the newer model (August 2025, against March 2025).
- Only Command A offers structured output.
Which should you choose?
Each answer below comes straight from the published prices, specifications and independent benchmark results.
Latest release
Choose Codestral
Codestral is the newer release (August 2025, against March 2025 for Command A).
Specifications side by side
| Codestral | Command A | |
|---|---|---|
| Provider | Mistral AI | Cohere |
| Category | Language models | Language models |
| Status | Generally available | Generally available |
| Headline price | No data | No data |
| Input | No data | No data |
| Output | No data | No data |
| Cached input | No data | No data |
| Context window | 128k | 256k |
| Max output | No data | No data |
| Released | August 2025 | March 2025 |
| Knowledge cutoff | No data | No data |
| Licence | Proprietary | Proprietary |
| Capabilities | Tool calling | Tool calling, Structured output |
| API model id | codestral-2508 | command-a-03-2025 |
Benchmarks side by side
Leader in bold.
| Benchmark | Codestral | Command A |
|---|---|---|
| Terminal-Bench Hard | No data | 0.8% |
| GPQA Diamond | No data | 52.7% |
| Humanity's Last Exam | No data | 4% |
| IFBench (instruction following) | No data | 36.5% |
| τ²-Bench (agentic tool use) | No data | 15.2% |
| Long-context reasoning (AA-LCR) | No data | 21.3% |
Independent evaluations by Artificial Analysis, run the same way for every model, at the reasoning setting shown. Percentages unless marked as an index.
Frequently asked questions
Should I choose Codestral or Command A?
Choose Codestral for the latest release. Choose Command A for very long inputs.
Which is cheaper, Codestral or Command A?
A price comparison is not possible: one of them has no comparable published price.
Which has the larger context window, Codestral or Command A?
Command A reads more per request: 256k tokens against 128k.
What is the difference between Codestral and Command A?
Codestral is made by Mistral AI and Command A by Cohere. A price comparison is not possible: one of them has no comparable published price. Command A reads more per request: 256k tokens against 128k. Codestral is the newer model (August 2025, against March 2025). Only Command A offers structured output.
Which is newer, Codestral or Command A?
Codestral was released on August 2025 and Command A on March 2025.