Wan2.1 Upscaler + FrameGen — v3.0
Added Image compression before wan processing which should reduce vram consumption;
Runtime profile
- Access
- External workflow
- Status
- Listed
- Version
- wan2.1_upscale+frame-gen_v3.json
- Source
- Civitai
- Updated
- Nov 3, 2026
Source description
Removed ComfyUI-Essentials node;
Added Image compression before wan processing which should reduce vram consumption;
Added subgraphs to to get rid of the clutter;
A group has been added to simplify the selection of scaling and interpolation ratio.
What you need
- Model files
- 6.6 GiB known subtotal
Remaining hardware or download-size estimates are not available yet.
Requirements
Models
6- Not resolved
Diffusion model · Hugging Face · Kijai/WanVideo_comfy
Upscaler · 64.0 MB · PTH · Unknown
rife47.pth
Repository matchedCheckpoint · Unknown
Text encoder · 6.27 GB · SAFETENSORS · U8 + FP16 + FP32 + FP8 (tensor header) · Hugging Face · chatpig/encoder
VAE · 242 MB · SAFETENSORS · BF16 (tensor header) · Hugging Face · Comfy-Org/Wan_2.1_ComfyUI_repackaged
LoRA · Hugging Face · Kijai/WanVideo_comfy
Node packs
5comfyui-frame-interpolation
FoundNode pack · Registry
- Found
Node pack · Registry
- Found
Node pack · Registry
- Found
Node pack · Unknown
VHS_VideoInfo
PossibleNode pack · Unknown
AI-generated commentary
AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.
Purpose
Wan2.1 Upscaler + FrameGen v3.0 is a workflow for upscaling a video and generating interpolated frames; it takes a video as input and combines the processed frames into a video output.
Flow
Its graph loads a video, resizes and upscales its images, applies RIFE VFI, and combines the frames into a video.
Inputs and outputs
Provide a source video; the graph combines its processed frames into a video output.
Setup
The listed base model is Wan Video 1.3B t2v.
Dependencies
The listed node packs are ComfyUI-KJNodes (1.1.3), ComfyUI-Frame-Interpolation (1.0.7), and ComfyUI-VideoHelperSuite (4c7858ddd5126f7293dc3c9f6e0fc4c263cde079).
The listed model files are KijaiWan21T2V13B.safetensors, RealESRGAN_x2plus.pth, rife47.pth, umt5_xxl_fp8_e4m3fn_scaled.safetensors, wan_2.1_vae.safetensors, and Wan21_CausVid_bidirect2_T2V_1_3B_lora_rank32.safetensors.
The workflow also lists SimpleCalculatorKJ and VHS_VideoInfo, but gives no node-pack source for either.
Configuration
A control group lets you choose the upscale and interpolation ratios.
Version 3.0 adds image compression before Wan processing, with the stated aim of reducing VRAM use.
Cautions
The creator reports 160–480 seconds to scale and generate frames for a 33-frame video after the models load; the time varied with resolution.
Does this need to be edited?
Sign in to send an edit request.
Sources
1 source
Sources
1 sourceWorkflow instructions
Author notes from wan2.1_upscale+frame-gen_v3.json · selected release
Select upscale and interpolation ratio above.
Download files
Model artifacts
2 artifacts
Model artifacts
2 artifactsSource excerpts
3 excerpts
Source excerpts
3 excerptsSource context: 656 downloads · Type Workflows · Base model Wan Video 1.3B t2v
Source context: 630 downloads · Type Workflows · Base model Wan Video 1.3B t2v
Source context: 546 downloads · Type Workflows · Base model Wan Video 1.3B t2v
