Skip to content
WorkflowListedExternal
Open source

VACE Outpainting! Native & Wrapper Workflows + Video Guide

In this tutorial, I show you how to outpaint entire video sequences using the powerful VACE 14B model inside ComfyUI. This means expanding your AI-generated video frames beyond their original borders—adding...

Sourcecivitai.com/models/1636498
TaskMultiCapabilitiesImage edit · Text to videoFamilyWanEstimated VRAMNot yet characterizedBgrade

Runtime profile

Access
External workflow
Status
Listed
Source
Civitai
Updated
Oct 25, 2026

Source description

In this tutorial, I show you how to outpaint entire video sequences using the powerful VACE 14B model inside ComfyUI. This means expanding your AI-generated video frames beyond their original borders—adding backgrounds, scenery, and dynamic content that feels like it was always there. Whether you're trying to improve aspect ratios, enhance your shots, or add cinematic framing to your AI clips, this guide will show you how to outpaint with precision and style. Perfect for creators working with Wan2.1 or CausVid pipelines looking to push the limits of AI video quality and control.

Native & Wrapper Workflow Demos and Video Tutorial:

What you need

Model files
0 GiB known subtotal

Remaining hardware or download-size estimates are not available yet.

Requirements

Models

4

Node packs

2

AI-generated commentary

AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.

Purpose

  • This ComfyUI workflow is for outpainting AI-generated video sequences with VACE 14B: it takes an AI-generated video sequence and produces video with frames expanded beyond the original borders, adding backgrounds, scenery, or dynamic content.

Flow

  • The workflow loads a video, pads and resizes images, samples with Wan VACE, decodes frames, trims latent video, and combines the result into a video.

  • It includes text encoding and loaders for the text encoder, VAE, U-Net, and LoRA components.

Inputs and outputs

  • Input: an AI-generated video sequence. Output: a video with frames expanded beyond the original borders and added backgrounds, scenery, or dynamic content.

Setup

  • Use it inside ComfyUI; its listed base model is Wan Video 14B t2v.

  • The listing includes native and wrapper workflow demos plus a video tutorial.

Dependencies

  • The workflow lists four model files: native_umt5_xxl_fp8_e4m3fn_scaled.safetensors, native_wan_2.1_vae.safetensors, wan2.1_vace_14B_fp16.safetensors, and Wan21_CausVid_14B_T2V_lora_rank32.safetensors.

  • It requires the node packs comfyui-kjnodes and comfyui-videohelpersuite.

Configuration

  • Consider it when you want to improve an AI video’s aspect ratio, enhance a shot, or add cinematic framing.

    Suggestion · not verified

Cautions

  • Before using it, verify that these exact model filenames are available: native_umt5_xxl_fp8_e4m3fn_scaled.safetensors, native_wan_2.1_vae.safetensors, wan2.1_vace_14B_fp16.safetensors, and Wan21_CausVid_14B_T2V_lora_rank32.safetensors.

    Suggestion · not verified

  • Also verify that comfyui-kjnodes and comfyui-videohelpersuite are available in ComfyUI.

    Suggestion · not verified

Does this need to be edited?

Sign in to send an edit request.

Sources

1 source
CivitaiPrimary
civitai.com/models/1636498

Verified Jul 19

Source excerpts

1 excerpt
NoteJul 19civitai.com/models/1636498

Source context: 448 downloads · Type Workflows · Base model Wan Video 14B t2v