Skip to content
localmodel.run

Video model · wan

WA Wan 2.2 TI2V 5B requirements

DIT video model · 5B params · 1280×704 (720p), 121f (~5s) · released Jul 2025. Realistic minimum to run: Nvidia GeForce RTX 3060 (12GB) at Q4 GGUF.

Apache-2.0 Commercial use OK
Peak VRAM (Q4 GGUF)
~8 GB
All resident
~24 GB
Offload floor
~5 GB
Clip
121f / ~5s

Backbone size by precision

PrecisionOn disk
PrecisionSize
fp16 / bf16 10 GB
Q8 GGUF 5.4 GB
Q4 GGUF (recommended) 3.43 GB
Q2 GGUF 1.85 GB

Backbone weights only. Peak VRAM is dominated by the activation memory for 121 frames at 1280×704 (720p), not the file size.

Pipeline components

ComponentSize
ComponentSize
umT5-XXL text encoder offloaded 3.66 GB
VAE (3D) 2.82 GB

Video VAEs are larger than image VAEs because they decode a temporal stack of frames.

Run it

Wan 2.2 TI2V 5B runs in ComfyUI or Diffusers. Generating more frames or higher resolution raises peak VRAM sharply; the Q4 GGUF figure is for the default 121-frame clip.

ComfyUIDiffusers

Which devices can run Wan 2.2 TI2V 5B?

FAQ

How much VRAM does Wan 2.2 TI2V 5B need?

At Q4 GGUF the realistic peak is ~8 GB, versus ~24 GB with every component resident. With aggressive CPU offload it drops to ~5 GB, much slower.

Why is peak VRAM lower than the sum of the files?

The text encoder is run once to encode your prompt, then offloaded to CPU before the frames are generated, so it is not resident at the memory peak.

Can I use Wan 2.2 TI2V 5B commercially?

Yes. Wan 2.2 TI2V 5B is licensed Apache-2.0, which permits commercial use.

Wan 2.2's unified 5B text-and-image-to-video model (720p, 121 frames at 24fps). The official spec asks for a 24GB GPU with offload, but ComfyUI native offloading fits it well on 8GB, and GGUF Q4_K_M (3.43GB) + T5 on CPU lands around 5-8GB. Note the larger 2.82GB Wan 2.2 VAE. Apache-2.0, commercial OK. Anchor is the ComfyUI 8GB path (synthesis). Sources: Wan-AI card, QuantStack GGUF, ComfyUI Wan 2.2 tutorial, Wan2.2 repo.

Sources

VRAM is a sourced peak-usage anchor at Q4 GGUF (composed from component sizes, not a single measurement) for the default clip length, validated 2026-06-15. See methodology.