OpenShortsAn open source video clipper you can self-host
OpenShorts is an MIT-licensed video clipper that runs entirely on your own hardware via Docker Compose. Source video never leaves the machine.
The stack is Python 3.11, FastAPI, faster-whisper, PySceneDetect, MediaPipe, YOLOv8, FFmpeg and Google Gemini 3.0 Flash, with a React dashboard.
It is the only open source tool in this category. Opus Clip, Klap, Vizard and Submagic are all closed-source cloud services.
Why self-host a video clipper at all?
Three reasons come up repeatedly. The first is that unreleased footage, client work and internal recordings should not be uploaded to a third party whose retention policy you have not read. The second is cost at volume: a per-minute cloud tool gets expensive quickly if you process long recordings every week, whereas self-hosting costs electricity. The third is that the output is opinionated, and if you disagree with how it reframes or where it cuts, having the source means you can change it rather than file a feature request.
What is in the pipeline?
Transcription with word-level timestamps
faster-whisper transcribes the source audio and returns a timestamp for every individual word, not just every sentence. Word-level timing is what makes both the clip boundaries and the burned-in subtitles land on the right frame.
Scene boundary detection
PySceneDetect scans the video for cuts and hard visual transitions. These boundaries stop a clip from starting mid-shot, which is the most common reason an auto-generated clip reads as machine-made.
Viral moment scoring
Google Gemini 3.0 Flash receives the timestamped transcript and the scene boundaries together, and returns the 3 to 15 strongest segments of 15 to 60 seconds. Each one is scored on hook strength, emotional payload and how well it stands alone without the surrounding context.
Vertical reframing
Each clip is cropped from 16:9 to 9:16 in one of two modes. TRACK mode follows a single subject with MediaPipe face detection and a YOLOv8 fallback, damped by a stabiliser that holds the camera still inside a safe zone instead of chasing every head movement. GENERAL mode handles group shots and landscapes by keeping the full width over a blurred backdrop.
Subtitles, hooks and effects
Subtitles are generated from the word-level transcript and burned in with FFmpeg. An AI-written hook overlay covers the first seconds. Optional Gemini-generated FFmpeg filter chains handle colour and transitions.
Dubbing and publishing
ElevenLabs dubbing translates the audio into 30+ languages while preserving the speaker's voice, and the dubbed track is re-transcribed so the subtitles match the new language. Finished clips post directly to TikTok, Instagram Reels and YouTube Shorts.
What does it run on?
Docker Compose brings up the FastAPI backend and the React dashboard together. The realistic floor is 8GB of RAM and a modern multi-core CPU; an NVIDIA GPU is optional and takes an 8-minute video from roughly 5 to 8 minutes of processing down to about 50 seconds. Linux, macOS and Windows via WSL2 are all supported. Concurrency is controlled by a semaphore configured with MAX_CONCURRENT_JOBS.
What is the licence?
MIT for the core application, which means you can use it commercially, modify
it and redistribute it. The cloud/ directory, which contains
billing, managed keys and the hosted-service infrastructure, is carved out under
a separate commercial licence and is not needed to self-host.
How does it compare to the closed-source tools?
OpenShorts is the only open source option in this category. As of July 2026, Opus Clip starts at $15/month, Submagic from $14/month, Vizard at $19.99/month and Klap at $29/month, and none of them can be self-hosted or audited. The trade-off is real in both directions: they ship more caption presets and require no setup, and you cannot read a line of what they do with your video.
Common questions
- Is there an open source alternative to Opus Clip?
- Yes. OpenShorts is MIT-licensed and self-hostable with Docker, and covers the same core job: AI moment detection, face-tracked 9:16 reframing and word-level subtitles. Opus Clip is closed source and cloud only, starting at $15/month.
- Can I run it without sending video to any third party?
- Transcription, scene detection, reframing and encoding all run locally. Moment scoring calls the Google Gemini API, which receives the transcript rather than the video file. Dubbing and AI UGC generation are optional and call ElevenLabs and fal.ai respectively; leave them off and nothing but transcript text leaves the machine.
- What licence is OpenShorts released under?
- MIT for the core application. The cloud/ directory covering billing and hosted infrastructure is under a separate commercial licence and is not required for self-hosting.