LLM Memory Calculator

How much VRAM does a model need? Estimate memory requirements at every quantization level — FP32 down to 1.58-bit — and check hardware fit instantly.

Runs entirely in your browser. Nothing is sent anywhere.

Model parameters

Enter model size or select a known architecture. Overhead adds ~10–20% to the raw weight footprint for KV cache, activations, and framework buffers.

Quick select

Memory by quantization

Format Bits/param Raw (GB) With overhead Hardware fit
Paper-ready description
Hardware reference
GPU / SetupVRAMRecommended max load
RTX 306012 GB~10 GB
RTX 409024 GB~20 GB
Kaggle T4 ×116 GB~13 GB
Kaggle T4 ×232 GB~26 GB
A100 40 GB40 GB~34 GB
A100 80 GB80 GB~68 GB
CPU only (RAM)variesdepends on system RAM