Well-Made

ComfyUI nodes for Wan 2.2 SVI 2 Pro with Keyframe control via First/Last Frame and seamless video stitching.

37
8
100% credibility
Found Feb 26, 2026 at 19 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Custom nodes for ComfyUI that combine motion continuity from one video model with first/last frame control from another to generate smooth, stitched videos from image sequences.

How It Works

1
🔍 Discover smooth video magic

You hear about a handy add-on that lets you create flowing videos by guiding each part with a starting picture and an ending picture.

2
📥 Add it to your creative space

Easily install it into your video-making workspace with a quick search or simple download and restart.

3
📋 Load ready-made plans

Open one of the example setups that show how to build videos from pictures, ready for up to several connected parts.

4
🖼️ Choose your guide pictures

Pick a starting image to anchor the beginning and an optional ending image to shape how each video segment finishes.

5
Generate flowing segments

Watch as your workspace creates smooth video clips that blend motion perfectly from one part to the next.

6
🔗 Connect and blend parts

Link the segments together, trimming ends if needed for seamless flow without jumps or glitches.

7
⚙️ Fine-tune for perfection

Adjust sliders for overlap or brightness to make transitions crystal clear and colors vibrant.

🎉 Enjoy your masterpiece

Sit back and play your long, smooth video that tells a continuous story just as you imagined.

Sign up to see the full architecture

6 more

Sign Up Free

Star Growth

See how this repo grew from 19 to 37 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is ComfyUI-Wan-SVI2Pro-FLF?

This Python-based ComfyUI custom nodes pack delivers precise first/last frame control for Wan 2.2 SVI 2 Pro video generation, solving choppy multi-segment stitching by blending motion continuity with hard-locked end frames. Users get seamless video extension from images or existing clips—up to 7 segments out of the box, scalable by duplicating workflows—via included JSON examples for KSampler or SamplerCustomAdvanced. Install via ComfyUI custom nodes manager, git clone from the ComfyUI custom nodes GitHub repo, or ComfyUI GitHub releases, with nodes landing in conditioning/video_models and latent/video categories.

Why is it gaining traction?

It stands out by merging SVI 2 Pro's smooth motion tails with FLF-style end anchoring, dodging ghosting at seams through overlap tweaks and latent trimming—features absent in stock KJNodes or native Wan setups. Developers dig the optional end_samples input for flexible prompt-only or frame-guided extensions, plus fixes for ignored last frames and washed-out outputs via built-in contrast adjustments. Workflows handle ComfyUI custom nodes conflicts gracefully, with ComfyUI GitHub examples showing real results on AMD, Mac, or portable installs.

Who should use this?

ComfyUI video artists chaining 81-101 frame segments for long-form AI clips, especially those tweaking Wan 2.2 models from CivitAI or Hugging Face. Ideal for NSFW/SFW creators needing deterministic endings without rephrasing prompts endlessly, or hobbyists extending videos via ComfyUI custom workflows and ComfyUI custom scripts.

Verdict

Grab it if you're deep in ComfyUI video pipelines—solid docs and workflows punch above its 19 stars and 1.0% credibility score, though low adoption signals early maturity; test on your model combo first to sidestep VAE seam quirks. Skip for single-clip work; it's niche but effective for stitching pros.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.