Command A Vision
active
Cohere2025-07-31
Specifications
Modalities
text
image
Context Window128K tokens
Max Output8K tokens
Training Cutoff-
Open SourceNo
Pricing (per 1M tokens)
Input$-
Output$-
Benchmarks
MMLU-Pro-
SWE-Bench-
GPQA Diamond-
HumanEval-
ARC-AGI-
MATH-
API
Model ID
command-a-vision-07-2025Sources
- Cohere: An Overview of Cohere's Models
general · as of 2026-07-04
- Cohere: Cohere's Command A Vision Model
release date