Home › AI glossary › VRAM
What is VRAM?
VRAM is the fast memory on a graphics card, and it decides how large an AI model you can run on your own computer.
- Category Hardware
- Also called video memory
- Updated 20 Oct 2026
VRAM, explained simply
VRAM stands for video random-access memory. It is the memory that lives on the graphics card, next to the GPU.
An AI model must fit into VRAM to run quickly. A bigger model needs more VRAM.
If a model is too big, it may run very slowly or not start. Many guides tell you how much VRAM a model needs.
On ITG Now, our free tool guides mention how much memory a model needs when it runs on your own computer.
Related terms
GPURAMLocal AIQuantizationParametersModel weights
Where you'll see it on ITG Now
Questions people ask about VRAM
VRAM vs RAM: what is the difference?
RAM is the main memory the whole computer uses. VRAM sits on the graphics card and feeds the GPU directly.
How much VRAM do I need for AI?
It depends on the model. Bigger models need more, and smaller or compressed ones need less. Check the tool's guide.
Can I add more VRAM to my computer?
Usually not. VRAM is part of the graphics card, so you would need a different card.