GenModelHub

Concat Video Editor: What It Is and How to Join AI Video Clips (2026)

Updated 2026-10-19

Concat is a free, open-source video editor that runs on your own computer or phone and is pitched as a CapCut replacement; its GitHub page calls it "the truly free, open-source, local-first, cross-platform CapCut replacement". The project (jub0t/Concat on GitHub) is licensed AGPL-3.0-or-later, shows about 9.3k stars, and its latest listed release on 19 October 2026 is version 0.2.8, which the project itself labels beta. AI video models usually return 4–15 second clips, so joining several of them is the normal next step – and a local editor like Concat does that without uploading your clips anywhere. GenModelHub is independent and not affiliated with Concat; GenModelHub makes the clips, it does not join them.

What Concat does (per its GitHub page)

Feature What the page says
Platforms Windows (x86_64), macOS (Intel and Apple silicon), Linux (x86_64 and ARM), Android; iOS/iPadOS by sideloading
Editing multi-track timeline, titles, keyframe animation, about 100 effects, filters and transitions
AI-assisted auto-captions with local Whisper, text-to-speech and voice cloning, background removal, voice cleanup
Export H.264, HEVC and AV1, up to 4K at 60 fps with 10-bit colour
Automation built-in MCP server (concat --mcp) and a JSON-RPC API for scripts
Price / licence free; AGPL-3.0-or-later; beta

Build-from-source needs Rust 1.93+, FFmpeg 7+ and cmake; the prebuilt downloads are on the project's releases page. Because it is beta, expect rough edges and check the project's issues before relying on it for a deadline.

How to join several AI video clips into one video

The steps are the same in Concat or any timeline editor; they are written for Concat but not verified against every menu name.

  1. Generate the clips first, one shot each. Keep the same aspect ratio (for example 9:16) and the same model for every clip so colour and motion match. On GenModelHub the clip lengths are 4–15 seconds depending on the model.
  2. Download the MP4s and put them in one folder, named in the order you want (01, 02, 03).
  3. Drag them onto the timeline in that order. Put a short transition or a straight cut between them; a cut on a moving action looks better than a fade.
  4. Add the sound last. Many clips come without sound, or each clip has different ambient sound. One music or voice track across the whole timeline hides the seams.
  5. Export in H.264 for TikTok, Reels and Shorts.

Make clips that cut together

  • End each shot on a clear action or a held pose, so the next shot has a natural entry point.
  • Reuse the same reference photo for the same character in each image-to-video clip (see image to video); faces drift between separate text-only generations.
  • Use the same lighting words in every prompt (for example "warm evening light").
  • Test short. The cheapest GenModelHub clips cost 4–10 credits, roughly $1–3 each, and you see the cost before generating (one credit is about $0.28–0.30). See the cheap AI video generator page for the model list. There are no free credits and failed generations are refunded.

Concat vs other ways to join clips

Option Cost Runs where Note
Concat free, open source your device beta; full editor
CapCut free tier plus paid plans app / web the editor Concat positions itself against
FFmpeg concat free command line no timeline; clips must share codec and size
The AI tool's own extend varies in the tool stays inside one model's style

FAQ

What is the Concat video editor? A free, open-source, local-first video editor for Windows, macOS, Linux and Android, described by its authors as a CapCut replacement.

Is Concat free and open source? Yes. Per its GitHub page it is AGPL-3.0-or-later licensed and free to download.

How do I join several AI video clips into one video? Generate each shot as its own clip, then place them in order on an editor's timeline and export one MP4. Steps above.

Does GenModelHub join clips? No. It generates 4–15 second clips; join them in Concat, CapCut or any editor.

Is Concat ready for production work? It is labelled beta (version 0.2.8). Test your export before a deadline.

Source (19 Oct 2026): Concat on GitHub. Feature lists are the project's own claims, not independently tested.

Try the Image to Video AI

Upload a photo, describe the motion, and get a short video.

Generate a video