Short Video Maker
Open-source tool that generates short-form videos from text with text-to-speech, captions, Pexels background footage and music via REST API and MCP server.
At a Glance
Self-hosted short video maker (MCP and REST server) under the MIT License. Free alternative to GPU-heavy video generation and paid third-party APIs.
Engagement
Available On
Alternatives
Listed Nov 2026
About Short Video Maker
Short Video Maker is an open-source server published by gyoridavid that turns simple text inputs and search terms into short vertical or landscape videos for TikTok, Instagram Reels, and YouTube Shorts. It combines text-to-speech, automatic captions, background videos, and music, and exposes both a REST API and a Model Context Protocol (MCP) server. The repository was open-sourced by the AI Agents A-Z YouTube channel and is MIT licensed.
What It Is
Short Video Maker is a self-hosted video creation service. According to the README, it is meant as a free alternative to GPU-heavy video generation and to paid third-party API calls. It does not generate video from scratch from an image or an image prompt, and it cannot stitch user-supplied images or videos together.
Each video is built from scenes. A scene has narration text, which the text-to-speech engine reads and uses for captions, and search terms used to find background footage from Pexels. If no footage is found, fallback terms such as nature, globe, space, and ocean are used.
How the Pipeline Works
The workflow has five steps: Kokoro TTS converts text to speech, Whisper generates accurate captions, Pexels supplies relevant background videos, Remotion composes the elements, and the server renders the finished short with timed captions. Configuration options cover end padding, background music mood, caption position and background colour, Kokoro voice, orientation (portrait or landscape), and music volume.
Interfaces and Automation
The REST API can create a video, check its status, download it, list or delete videos, and list available voices and music tags. The MCP server offers create-short-video and get-video-status tools for AI agents such as n8n; the README notes the REST endpoints give more flexibility. A community-built web UI is available at localhost:3123, and example n8n workflows and a tutorial video are linked.
Setup Path
Docker is the recommended route, with tiny, normal, and CUDA images that differ in Whisper and Kokoro model size and GPU acceleration. It can also run through npm or npx on Ubuntu 22.04 or newer and macOS; Windows is not supported at the moment. A free Pexels API key and internet access are required. Documented limitations include English-only voiceover and Pexels as the only background video source.
Community Discussions
Be the first to start a conversation about Short Video Maker
Share your experience with Short Video Maker, ask questions, or help others learn from your insights.
Pricing
Open Source (MIT)
Self-hosted short video maker (MCP and REST server) under the MIT License. Free alternative to GPU-heavy video generation and paid third-party APIs.
- MIT License for the project code
- Text-to-speech via Kokoro TTS
- Automatic captions via Whisper
- Background videos via Pexels (free Pexels API key required)
- REST API and MCP server
Capabilities
Key Features
- Generate complete short videos from text prompts
- Text-to-speech with Kokoro voices
- Automatic caption generation and styling
- Background video search and selection via Pexels
- Background music with mood selection
- REST API and Model Context Protocol (MCP) server
- Web UI at localhost:3123
- Portrait and landscape orientation
- Docker images for tiny, normal and CUDA setups
Integrations
Demo Video

