PyPI · AI & LLM
ollama-scout
Reads your hardware and tells you which local models will run.
About
A CLI that scans GPU VRAM, CPU and system RAM, then recommends Ollama models that fit, grouped by what you want to do: coding, chat, reasoning or vision. It saves the trial-and-error of pulling a 30GB model only to watch it run out of memory.
What it does
- Detects NVIDIA, AMD and Apple Silicon memory, including unified memory
- Sizes recommendations by quantization and use case
- Guided interactive mode for beginners, plain flags for scripting
How it's built
Recommendations come from the memory the machine actually has, not from a model's advertised size.