Replace or remove an object in a video
Open this ready-to-run workflow on Floyo. No install needed.
HOW IT WORKS
Step 1. Upload your video. The clip you want to edit. The workflow loads up to 121 frames at 24fps, about 5 seconds. Shorter clips work too. Works great with: product shots · talking heads · street scenes · VFX plates
Step 2. Draw a mask. Paint over the region you want to replace or fill, a little larger than the target object. Everything outside the mask stays untouched.
Step 3. Describe what fills the mask. Write the full scene, not only the new object. "A woman walking through a park holding a red umbrella, warm afternoon light" beats "red umbrella," because the model needs context to match the lighting and perspective. Add a negative prompt if you want.
Step 4. Hit run and download. The model inpaints the masked region in a single pass. Preview it in the workflow, then download. Ready for: Premiere · DaVinci Resolve · After Effects · any editor
First time? Leave every setting as-is. The defaults (8 steps · guidance 1 · seed 42) are the right starting point. Put your effort into the mask and the prompt.
Overview
This workflow replaces, adds, or removes an object in a video using LTX Video 2.3 and a dedicated inpainting LoRA. You mask a region, describe what belongs there, and the model regenerates only that area while the rest of the footage stays exactly as shot. It matches the lighting and perspective of the surrounding frames, so the fill reads as part of the original shot, not a paste-over. This is the single-pass version, built for speed. It skips the upscale refinement passes, so you can iterate on mask placement and prompt wording fast, then move to the multi-pass version for final delivery. A run comes back in about 2 minutes. No setup, no nodes to wire.
Who it's for: editors, VFX artists, and product teams who want masked video editing in ComfyUI without wiring an inpainting pipeline from scratch. Not for: full-frame restyling or a mask that covers most of the shot. Plain video-to-video generation is the better tool once you are changing the whole frame.
Why Floyo
Floyo is the only ComfyUI platform built for teams in the browser.
Made for teams. Share run history, files, and models across your whole team. A teammate opens your exact run and picks up where you left off. No file handoffs, no version confusion.
No install, no setup. Every workflow and model is preloaded. Open it in your browser and run. Nothing to download, nothing to configure.
No local hardware. Workflows run on H100 NVL GPUs, so heavy models run fast without a card of your own. Your VRAM stops being the limit.
Open and closed models in one place. Floyo runs open-source workflows and API models side by side.
How to use (in your browser on Floyo)
Upload your video and mask the region you want to change.
Describe the full scene including the new object, then run.
Download the edited clip, ready for any editor.
Expectations The models are preloaded, so there is nothing to download. Running it needs a free Floyo account. Each run comes back in about 2 minutes. This is the fast, single-pass version, so treat it as your iteration pass: test mask positions and prompt wording here, then switch to the multi-pass version for a sharper final. A few things decide the result. The prompt is the most important setting, so describe the whole scene, not just the masked object. Mask a little larger than the target, since the model needs room to blend. Locked-camera footage with steady lighting inpaints cleaner than fast-moving subjects that cross the mask edge. LTX Video 2.3 is Apache 2.0 licensed, so your edited clips are yours to use commercially, including client and broadcast work.
Use Cases
Object Replacement. Swap a prop, a logo, or an unwanted element and describe what should take its place, matched to the footage.
Object Removal. Mask wires, signage, or background clutter and prompt the model to fill with the surrounding scene.
Product Placement Previews. Drop a product into existing footage for quick review before a full-quality render.
Fast Iteration. Test mask and prompt variations in the time a multi-pass render takes to finish one.
What can LTX 2.3 video inpainting change?
