Earngenix Logo
Skip to main content

Glossary · Models & Files

What Is fp8 in ComfyUI?

By Earngenix Team ·

⚡ Quick Answer

fp8 is a way of storing a model's numbers using 8 bits of precision instead of the usual 16. Think of it like rounding a long decimal to fewer digits — you lose a tiny bit of exactness, but the file gets roughly half the size and needs roughly half the VRAM to load.

Because the quality loss is usually small, fp8 has become a common "middle ground" release alongside a checkpoint's full-precision version.

Where You'll See It

Model download pages often list an "fp8" version next to the full bf16 or fp16 release. Some ComfyUI loader nodes also have a weight_dtype setting where you can pick an fp8 option (like fp8_e4m3fn) even when loading a full-precision file.

Quick Example

A checkpoint's bf16 version might need around 24 GB of VRAM to load. Its fp8 version needs roughly half that, letting it fit comfortably on a 12 GB card with output that looks very close to the original.

Newer NVIDIA GPUs (recent Ada and Blackwell generations) have hardware built specifically to accelerate fp8 math, so they benefit from both the VRAM savings and a real speed boost. Older cards mainly get the VRAM savings.

Frequently Asked Questions

You can load an fp8 model on almost any GPU to save VRAM, but newer NVIDIA GPUs (Ada and Blackwell generations) also get a speed boost from hardware that natively accelerates fp8 math, not just older cards.

No. fp8 is one specific, fixed precision level, while GGUF is a whole file format offering several different quantization levels to pick from. Both shrink VRAM use, just in different ways.

Generally yes — most LoRA loaders apply the LoRA on top of the model regardless of which precision it was loaded at, though results can vary slightly by node and model family.

See It In Action

Not sure which version to download?

Our VRAM & Hardware Guide matches models to what your GPU can actually run.

Published: 2026-09-17 · Last updated: 2026-09-17

Discussion

Join the discussion

Sign in to leave a comment or reply

💬

No comments yet

Be the first to share your thoughts!