Run Command A locally
Command A is 111B parameters (dense). Here's the VRAM it needs at each quantization, the smallest GPU that fits, and how a self-hosted box compares with the cheapest verified API.
VRAM by quantization
| Quantization | VRAM needed | Fits on | Notes |
|---|---|---|---|
| int4 (Q4) | 66.6 GB | A100 (80 GB) ($1.19/hr) | smallest footprint, minor quality loss |
| int8 (Q8) | 133.2 GB | 2× H100 (160 GB) ($3.98/hr) | near-lossless |
| fp16 (full) | 266.4 GB | 4× H100 (320 GB) ($7.96/hr) | reference quality |
VRAM ≈ params × bytes/param × 1.2 (overhead). See the fullmethod and break-even calculator.
Self-host vs API
Cheapest API for Command A is $12.5 / 1M tokens(blended) via Cohere. Running it yourself in int4 fits aA100 (80 GB) at $1.19/hr — about $868.7/month at 24/7. Those two lines cross at roughly 69.5M tokens/month: below that the API wins on cost, above it the dedicated GPU does (assuming you keep it busy). Tune your own volume in the break-even calculator.
Run it at home
No cloud account needed: Command A in int4 (66.6 GB) fits a Mac Studio M5 Ultra (96 GB unified), a $5,499 one-off buy. Amortized over 3 years that's about $0.2092/hr — hardware only, electricity aside. Against the Cohere API at $12.5/1M, buying it pays for itself after roughly 440M tokens total. Below that the API is cheaper; a machine that mostly sits idle rarely earns back its price.
Get the weights
Quantized builds on the model hubs — links search live, so they track new builds as they appear:
- GGUF builds llama.cpp · Ollama · LM Studio
- AWQ builds GPU serving (vLLM)
- GPTQ builds GPU serving
- MLX builds Apple silicon
- Ollama library one-command local run