Why we are featuring it
Choose a model based on the machine you really have.
whichllm detects GPU, CPU, and RAM, then ranks Hugging Face models that appear suitable for the system. It can simulate hardware configurations and distinguish what fits from what runs well.
What is inside
- Detects hardware and ranks compatible models.
- Simulates GPUs and multi-GPU workstations.
- Filters by VRAM, speed, and headroom.
- CLI, Markdown, JSON, and Python snippet output.