LLMs for Your GPU — Which Open Models Run on Your Graphics Card?

Pick your graphics card and see exactly which open-weight LLMs run on it — with VRAM fit, estimated tok/s and community speed reports for every model.

Open the Hardware Compass

Every fit guide is computed live from each model's parameter size, the chosen quantization level and your GPU's VRAM: a model fits when its quantized weights plus KV-cache overhead stay within memory. Speed estimates combine memory bandwidth with real-world tok/s reports from the community, so you see both what runs and how fast it runs.