Wan2.1 T2V 1.3B SelfForcing VACE needs about 2.1 GB of memory at Q4_K_M (a 1.5 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 15 quants, from 0.8 GB (Q2_K) to 4.3 GB (F16).
Wan2.1 T2V 1.3B SelfForcing VACE – 2.15B parameter video diffusion model
Wan2.1 T2V 1.3B SelfForcing VACE is a 2.15B parameter autoregressive video diffusion model. It is trained with Self Forcing to match inference distribution and supports real‑time video generation on high‑end GPUs. The model is released under cc‑by‑nc‑sa‑4.0.
Good at
- Real‑time video generation
- Autoregressive rollout with KV caching
- High quality streaming on RTX 4090
Limits
- Requires GPU with at least 24 GB VRAM
- No vision support in the text prompt only version
- License restricts commercial use
Best for
- Generating short video clips from detailed prompts
- Streaming video content on powerful GPUs
- Research on autoregressive diffusion methods
Who it’s for. Researchers and developers needing a high‑performance video diffusion model for real‑time generation. The model is suitable for non‑commercial projects due to its cc‑by‑nc‑sa‑4.0 license.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 0.8 GB | 1.4 GB | Fits | Get file → | |
| Q3_K_S | 1.0 GB | 1.6 GB | Fits | Get file → | |
| Q3_K_M | 1.1 GB | 1.7 GB | Fits | Get file → | |
| Q3_K_L | 1.2 GB | 1.8 GB | Fits | Get file → | |
| Q4_0 | 1.3 GB | 2.0 GB | Fits | Get file → | |
| Q4_K_S | 1.3 GB | 2.0 GB | Fits | Get file → | |
| Q4_1 | 1.4 GB | 2.1 GB | Fits | Get file → | |
| Q4_K_M | 1.5 GB | 2.1 GB | Fits | Get file → | |
| Q5_K_S | 1.5 GB | 2.2 GB | Fits | Get file → | |
| Q5_0 | 1.6 GB | 2.2 GB | Fits | Get file → | |
| Q5_K_M | 1.6 GB | 2.3 GB | Fits | Get file → | |
| Q5_1 | 1.7 GB | 2.4 GB | Fits | Get file → | |
| Q6_K | 1.8 GB | 2.5 GB | Fits | Get file → | |
| Q8_0 | 2.3 GB | 3.1 GB | Fits | Get file → | |
| F16 | 4.3 GB | 5.3 GB | Fits | Get file → |
For 8 GB VRAM, use Q4_0 (file size 1.3 GB, estimated 2 GB memory). For 12 GB VRAM, Q5_0 (1.56 GB, 2.2 GB memory) is suitable. For 24 GB VRAM, the FP16 (4.31 GB, 5.3 GB memory) provides the highest quality.
What GPU do you need for Wan2.1 T2V 1.3B SelfForcing VACE?
These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is Q8_0 (2.3 GB file, about 3.1 GB in use).
- 6 GB VRAM: the best quant that fits is F16 (4.3 GB file, about 5.3 GB in use).
Which Wan2.1 T2V 1.3B SelfForcing VACE quant should you choose?
Q4_K_M is the usual starting point for Wan2.1 T2V 1.3B SelfForcing VACE: it balances file size and quality well. Here it is a 1.5 GB file.
The Q8_0 file is 2.3 GB, about 1.6 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q2_K at 0.8 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Wan2.1 T2V 1.3B SelfForcing VACE
How much VRAM does Wan2.1 T2V 1.3B SelfForcing VACE need?
About 2.1 GB at Q4_K_M with an 8K context window. That is the 1.5 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Wan2.1 T2V 1.3B SelfForcing VACE on an 8 GB GPU?
Yes, with F16: the 4.3 GB file needs about 5.3 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Wan2.1 T2V 1.3B SelfForcing VACE GGUF?
It depends on the quant: Q2_K is 0.8 GB, Q3_K_S is 1.0 GB, Q3_K_M is 1.1 GB, Q3_K_L is 1.2 GB, Q4_0 is 1.3 GB, Q4_K_S is 1.3 GB and 9 more.
Where can I download Wan2.1 T2V 1.3B SelfForcing VACE GGUF?
The files are on Hugging Face in the lym00/Wan2.1_T2V_1.3B_SelfForcing_VACE-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Wan2.1 T2V 1.3B SelfForcing VACE use?
Hugging Face lists the license as cc-by-nc-sa-4.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run this model on an RTX 4090 with 24 GB VRAM?
Yes, the model is designed to run on RTX 4090 and can generate real‑time video with the full FP16 or lower quantizations.
Is the model compatible with ComfyUI?
Yes, the GGUF files can be used in ComfyUI with the ComfyUI‑GGUF custom node.
Does the model support text prompts only?
The provided model is a text‑prompt version; it does not include vision input capabilities.