Command A Pro: pricing, context window and benchmarks
Command A Pro is a multimodal model from Cohere. It offers a 200,000-token context window, costs $14.28 per million input tokens and $68.05 per million output tokens, and averages 81.3 across the index's benchmark suite.
Pricing
Command A Pro costs $14.28 per million input tokens and $68.05 per million output tokens. A blended 3:1 input-to-output mix works out to about $27.72 per million tokens. Cached input is listed at $1.52 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $14.28 | per 1M tokens |
| Output | $68.05 | per 1M tokens |
| Cached input | $1.52 | per 1M tokens |
| Blended 3:1 | $27.72 | estimate per 1M tokens |
| 1M in + 1M out | $82.33 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 81.3 | 138 | 77th | |
| GPQA Diamond | 79.1 | 70 | 88th | |
| SWE-bench Verified | 71.2 | 90 | 85th | |
| HumanEval | 87.5 | 78 | 87th | |
| MATH-500 | 83.1 | 144 | 76th | |
| AIME 2025 | 78.2 | 112 | 81th | |
| LiveCodeBench | 74.0 | 109 | 82th | |
| IFEval | 89.0 | 82 | 86th | |
| TruthfulQA | 83.3 | 101 | 83th | |
| HellaSwag | 90.1 | 84 | 86th | |
| MMMU | 77.2 | 111 | 82th | |
| Arena Elo | 1434.6 | 108 | 82th |
What Command A Pro is good for
Command A Pro is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $14.28 per million input tokens, it sits in the mid-range of the index, and the 200,000-token context window is enough for large files and multi-file diffs.
Strengths
- Large 200,000-token context window for long documents and agent traces.
- Accepts more than text, including image input where the provider supports it.
- Coding benchmarks are a relative strength within its tier.
Trade-offs
- Proprietary licence: you cannot download the weights or self-host the model.
- Premium input pricing at $14.28 per million tokens.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Claude Haiku 4.5 Reasoning | Anthropic | 128000 | $15.11 | 81.4 |
| Command A Pro | Cohere | 200000 | $14.28 | 81.3 |
| Doubao 1.5 Pro Max | ByteDance | 400000 | $14.93 | 81.2 |
| Nova Pro Max | Amazon | 1000000 | $14.05 | 81.2 |
| Mistral Large 3 Instruct | Mistral AI | 400000 | $5.31 | 81.1 |
| Sonar Pro Max | Perplexity | 128000 | $14.01 | 81.1 |
FAQ
How much does Command A Pro cost?
Command A Pro is listed at $14.28 per million input tokens and $68.05 per million output tokens. Cached input is $1.52 per million tokens. A blended 3:1 input-to-output mix is about $27.72 per million tokens.
What is the context window of Command A Pro?
Command A Pro has a context window of 200,000 tokens and can emit up to 16,384 output tokens per request in this index.
Is Command A Pro open weights?
No. Command A Pro is a proprietary model with a proprietary licence, so you cannot download the weights.
How does Command A Pro score on benchmarks?
Command A Pro averages 81.3 across the scored benchmarks in this index, ranking #96 of 600 models. Its strongest relative result is Arena Elo at 1434.6.
What is Command A Pro best for?
Command A Pro is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $14.28 per million input tokens, it sits in the mid-range of the index, and the 200,000-token context window is enough for large files and multi-file diffs.