Very long inputs
Choose Command A Vision
Command A Vision reads up to 128k tokens per request, against 32k for Voxtral Small.
Choose Command A Vision for very long inputs and images as input. Choose Voxtral Small for self-hosting and data control.
Each answer below comes straight from the published prices, specifications and independent benchmark results.
Choose Command A Vision
Command A Vision reads up to 128k tokens per request, against 32k for Voxtral Small.
Choose Voxtral Small
Voxtral Small has open weights, so you can run it on your own servers; Command A Vision is only available through its provider.
Choose Command A Vision
Command A Vision accepts images as input; Voxtral Small does not.
| Command A Vision | Voxtral Small | |
|---|---|---|
| Provider | Cohere | Mistral AI |
| Category | Language models | Language models |
| Status | Generally available | Legacy |
| 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 | 32k |
| Max output | No data | No data |
| Released | July 2025 | July 2025 |
| Knowledge cutoff | No data | No data |
| Licence | Proprietary | Open weights |
| Capabilities | Vision | Audio input, Self-hostable |
| API model id | command-a-vision-07-2025 | voxtral-small-2507 |
Choose Command A Vision for very long inputs and images as input. Choose Voxtral Small for self-hosting and data control.
A price comparison is not possible: one of them has no comparable published price.
Command A Vision reads more per request: 128k tokens against 32k.
Command A Vision is made by Cohere and Voxtral Small by Mistral AI. A price comparison is not possible: one of them has no comparable published price. Command A Vision reads more per request: 128k tokens against 32k. Only Command A Vision offers vision. Only Voxtral Small offers audio input, self-hostable.