ShortGPT is an open-source AI framework for automating short-form video channels on YouTube Shorts and TikTok. It combines a language-model content engine with an automated video-editing engine, so a channel can go from an idea to a finished, captioned short with very little manual work. The project has earned about 8,100 stars on GitHub, and its most recent code updates arrived in early 2025.
The framework is organized around two engines. The content engine uses GPT-4-class models to research topics, write scripts optimized for short-form retention, and plan the structure of each video. The editing engine then assembles the video itself: it sources and cuts footage, generates voiceovers, adds subtitles and overlays, and renders the final clip in the vertical format the platforms expect. ShortGPT was one of the early open-source projects to demonstrate a fully automated “faceless channel” workflow, and it remains a popular reference implementation for creators experimenting with programmatic content. Being a framework rather than a hosted service, you run it on your own machine with your own API keys, and every step — script, voice, footage, captions — can be customized or swapped out.
ShortGPT is free and open source under the MIT license. Running it costs whatever you spend on the underlying model and media APIs you connect.
ShortGPT is for creators and developers who want to experiment with automated short-form video production and faceless channels using an open, hackable framework.