Model Size Calculator
Calculate the on-disk file size of a model from its parameter count and quantization precision, from FP32 down to 4-bit.
Last reviewed by the Radiatus Cloud team
Calculate the file size of a model from its parameters and precision.
Securing AI in production?
We build guardrails, governance & compliance for AI systems.
Calculate model file size
The file size of a neural network model is driven by its number of parameters and the precision in which each parameter is stored. This calculator computes the size from a parameter count in billions and the chosen precision: four bytes for 32-bit, two for 16-bit, one for 8-bit and half a byte for 4-bit. A seven-billion-parameter model saved in 16-bit precision is about fourteen gigabytes on disk.
The same model quantized to 4-bit shrinks to around three and a half gigabytes, a quarter of the 16-bit size.
Why model size matters
Knowing the on-disk size helps with storage planning, download time and deciding which quantization to use for distribution. Smaller quantized files are faster to download and load and fit more easily on limited storage, which is why quantized model files in formats designed for local inference are so popular. The size in gigabytes uses decimal units, while the gibibyte figure uses the binary units operating systems report.
File size closely tracks the memory needed to load the weights, so it is also a quick proxy for VRAM requirements. All calculation happens locally in your browser.
Related tools
- AI Prompt Leakage Analyzer — Paste a system prompt and a hostile user input to see whether the prompt holds secrets and whether the input carries injection patterns. Local, instant.
- LLM Data Exposure Checker — Check if text contains data likely to be memorized or exposed by LLMs.
- AI Usage Policy Generator — Generate an acceptable use policy for AI tools in your company.
- Model Hallucination Estimator — Estimate risk of hallucinations based on task type and temperature.
Frequently Asked Questions
How is model file size calculated?
It is the parameter count multiplied by the bytes per parameter, which the precision determines, giving the size of the stored weights.
How much does quantization shrink a model?
Each halving of the bit width halves the size. A 4-bit model is a quarter the size of the same model in 16-bit precision.
What is the difference between GB and GiB?
GB uses decimal units of a billion bytes, while GiB uses binary units of 1,073,741,824 bytes, which is how operating systems usually report sizes.
Does file size equal the VRAM needed?
It is close, since the weights dominate, but running the model also needs memory for activations, overhead and any key-value cache.
Privacy & Security
Everything runs in your browser; nothing is uploaded.
How to Use
Enter the parameter count in billions and choose the precision.
Disclaimer: This tool is provided "as is" without warranty of any kind. Results are for educational and utility purposes.
Related Tools
AI Prompt Leakage Analyzer
AI SecurityPaste a system prompt and a hostile user input to see whether the prompt holds secrets and whether the input carries injection patterns. Local, instant.
LLM Data Exposure Checker
AI SecurityCheck if text contains data likely to be memorized or exposed by LLMs.
AI Usage Policy Generator
AI SecurityGenerate an acceptable use policy for AI tools in your company.