Description from extension meta
The best VRAM Calculator for Hugging Face models
Image from store
Description from store
It can be pretty hard to figure out what hardware you need to run LLMs and other AI models.
Now, you can easily see the requirements to inference or fine-tune a model. If it doesn't fit, you will even get suggestions on how to still get it working (i.e. quantization or QLoRA).
This tool is an extension to the Hugging Face website. This is still a Beta version so please leave feedback if you can think of improvements.
Latest reviews
- (2025-05-06) David de Vries: Works great
- (2025-05-06) Tim Schönborn: Super useful to quickly estimate whether a model on huggingface fits on my available GPU. Really nice that you can also set context length and batch size parameters.
- (2025-05-06) Albert Negura: Incredible tool!
- (2025-05-06) Luigi Pagani: Amazing!
- (2025-05-06) Willem Momma: Wow amazing tool!