Memory fit

Will this model fit on your GPU?

Enter any Hugging Face model. The lab reads its architecture, adds up weights, KV cache and runtime overhead, and tells you whether it fits and how much context you can afford.