Back to wiki Inference

Video quantization

Quantization shrinks Wan weights so they fit a smaller GPU. Official generate.py only casts dtypes (--convert_model_dtype). True FP8, GGUF, distillation and cache live in LightX2V, DiffSynth, FastVideo, Cache-dit, and ComfyUI GGUF loaders.

FP8GGUFLightX2V
On this page 5
BF16, FP8, GGUF and distilled NVFP4 paths for Wan DiT
Four different graphs. Do not use “quantized” as if they were one format.

What the official repo does

--convert_model_dtype converts parameters to config.param_dtype (typically bf16/fp16). Combined with --offload_model True and --t5_cpu it is the supported 80 GB → slightly less path — not 4-bit.

Community compression

  • DiffSynth-Studio — FP8, layer-offload, sequence parallel, LoRA.
  • LightX2V — step distillation, quantized packs, lightweight VAE, sparse attention.
  • FastVideo — distilled Wan with sparse attention.
  • Cache-dit — DBCache / TaylorSeer / Cache-CFG on Wan2.2 MoE (skips compute, not bits).
  • ComfyUI-GGUF / Kijai WanVideoWrapper — GGUF and kernel-level tricks.

How to choose

Need official sampling quality → bf16 + offload or multi-GPU. Need a 4090 at near-full quality → FP8 (see FP8 vs GGUF). Need 12–16 GB → GGUF or a distilled LightX2V pack, and expect texture loss.

FAQ

Does quantization add audio?
No.
Is cache quantization?
No. Cache-dit reuses activations across steps. Weights stay full precision unless you also quantize.

Primary sources

Checked against Wan-Video/Wan2.2 README · Aug 28, 2026

© 2026 wan2.video