Cohere - Dense - verified
Run Cohere Command A (03-2025) locally
Cohere Command A (03-2025) is a 111.1B dense model from Cohere with a 262K-token context window, released under CC BY-NC 4.0 (Cohere Labs) (commercial use is not allowed); the smallest listed download is 62.5 GB, the 4bit MLX build from mlx-community/c4ai-command-a-03-2025-4bit.
Official weights: CohereLabs/c4ai-command-a-03-2025
- Total params
- 111.1B
- Architecture
- Dense
- Context
- 262K
- Smallest download
- 62.5 GB
Fits in
Smallest download at most 85% of memoryDownloads
The official weights and every quantization listed for this model, with file sizes from Hugging Face.
| Download | Format | Size | Repo |
|---|---|---|---|
Official weights | BF16 | 222.1 GB | CohereLabs/c4ai-command-a-03-2025 |
Q4_K_M | GGUF | 67.1 GB | bartowski/CohereForAI_c4ai-command-a-03-2025-GGUF |
Q8_0 | GGUF | 118.0 GB | bartowski/CohereForAI_c4ai-command-a-03-2025-GGUF |
4bitSmallest listed | MLX | 62.5 GB | mlx-community/c4ai-command-a-03-2025-4bit |
GB is decimal: 1 GB = 10^9 bytes. Add memory for the context and the operating system on top of the file size.
Run it
Copied verbatim from the page linked under each command. Server commands from model cards often assume a multi-GPU node; check the flags before you run them.
Not sure which runtime to use? Compare local LLM runtimes by platform, API and features.
License
- License
- CC BY-NC 4.0 (Cohere Labs)
cc-by-nc-4.0 - Commercial use
- Commercial use: NoCommercial use is not allowed.
- Gated download
- Yes. Hugging Face asks you to accept the terms before downloading.
Worth knowing
Gated non-commercial download; the card says the model supports 256K context but the Hugging Face config ships set to 128K.
Released (Hugging Face repo creation date).
FAQ
How big is the Cohere Command A (03-2025) download?
The official BF16 weights in CohereLabs/c4ai-command-a-03-2025 are 222.1 GB. The smallest listed download is 62.5 GB, the 4bit MLX build from mlx-community/c4ai-command-a-03-2025-4bit. Plan for more memory than the file size: the context (KV cache) and the operating system need room too.
Can I use Cohere Command A (03-2025) commercially?
Cohere Command A (03-2025) is released under CC BY-NC 4.0 (Cohere Labs), and commercial use is not allowed. The Hugging Face download is gated: you accept the license terms before it starts. The license text is at https://cohere.com/c4ai-cc-by-nc-license.
How much memory do I need for Cohere Command A (03-2025)?
By this directory's rule of thumb it fits in 128 GB and up: the smallest listed download, 62.5 GB, is at most 85% of 128 GB, which leaves room for the context and the operating system.
How do I run Cohere Command A (03-2025) with Ollama?
Run "ollama run command-a", as shown at https://ollama.com/library/command-a. The Ollama library lists the command-a:111b tag at 67GB.
What is the context length of Cohere Command A (03-2025)?
262,144 tokens. A longer context needs more memory for the KV cache, so you can run with a shorter one when memory is tight.
Sources
Last verified from the pages below.
- HF API (params, license, createdAt) - huggingface.co
- HF file tree - huggingface.co
- Model card (model page) - huggingface.co
- Ollama tags - ollama.com
- GGUF Q4_K_M tree (bartowski) - huggingface.co
- GGUF Q8_0 tree (bartowski) - huggingface.co
- MLX tree - huggingface.co
Newsletter
New open weights, every week
One email per week. New model releases, open-weight drops and what they mean for what you build.
Keep exploring
Best local AI models
Our curated picks by hardware tier, with benchmarks and the reasoning.
Model API pricing and specs
First-party prices per 1M tokens when you would rather call an API.
AI benchmark leaderboards
Sourced scores, for when you need to compare quality.
Setup guides
Point your coding tools at the model you picked.