Repolode
Submit
IliasHad

IliasHad/edit-mind

Self-hosted, local-first video knowledge base that indexes your videos with transcription, frame analysis and embeddings, then lets you search scenes in natural language.

Open-source alternative to:Google Photos

View on GitHubWebsite
Stars
1,813
Forks
123
Open issues
15
Last commit
4 months ago

What is it?

Edit Mind indexes a video library by running transcription, frame analysis and multi-model embeddings on each file. It detects objects, faces and text, and extracts scene-level metadata, so you can search by spoken words, objects or people and jump to the matching scenes.

The app runs entirely on your own machine or server. It is started with Docker Compose and combines a React Router web app, a Node.js background job service using BullMQ, a Python ML service with PyTorch and Whisper, ChromaDB for vector search, and PostgreSQL via Prisma. A background watcher picks up new video files in the folders you add.

For natural language queries and metadata enrichment, you can point it at a local Ollama model or the Gemini API. The project is in active development and is not yet production-ready, and an optional commercial desktop app for macOS and Windows is offered separately.

Key features

  • 01Transcribes spoken audio from videos with Whisper.
  • 02Detects objects, faces and text in video frames.
  • 03Searches videos and specific scenes using natural language via ChromaDB vector search.
  • 04Watches configured folders and queues new files for background processing.
  • 05Runs locally in Docker, with an optional CUDA compose file for NVIDIA GPUs.

Use cases

  • Finding a moment in a large personal or family video archive by what was said or who appears.
  • Locating b-roll or specific shots across footage libraries for video editing.
  • Self-hosting a private video search service without sending footage to cloud services.

Quick start

  1. 1

    Run the one-line installer script

    curl -sSL https://get.edit-mind.com | sh
  2. 2

    Start all services with Docker Compose

    docker compose up
  3. 3

    Open the web app in your browser and log in

Things to know

  • The project is in active development and is not yet production-ready.
  • It uses a custom Edit Mind License rather than a standard open-source license, so check LICENSE.md before use.
  • Setup needs Docker and either a local Ollama model or a Gemini API key.

Topics

#video-indexing#computer-vision#face-recognition#semantic-search#self-hosted#docker#whisper#vector-database#ai#deepface

FAQ

What is edit-mind?
Self-hosted, local-first video knowledge base that indexes your videos with transcription, frame analysis and embeddings, then lets you search scenes in natural language. Edit Mind indexes a video library by running transcription, frame analysis and multi-model embeddings on each file. It detects objects, faces and text, and extracts scene-level metadata, so you can search by spoken words, objects or people and jump to the matching scenes.
Is edit-mind free to use? What license does it have?
edit-mind is open source under the Other license. Check the license terms before using it in commercial products.
Is edit-mind still maintained?
Updated within 6 months. The last commit to edit-mind was on Jun 30, 2026.
What can I use edit-mind for?
Finding a moment in a large personal or family video archive by what was said or who appears. Locating b-roll or specific shots across footage libraries for video editing. Self-hosting a private video search service without sending footage to cloud services.
How do I install edit-mind?
Run the one-line installer script: curl -sSL https://get.edit-mind.com | sh Start all services with Docker Compose: docker compose up Open the web app in your browser and log in
Where can I get edit-mind?
The source code and installation instructions are on GitHub: https://github.com/IliasHad/edit-mind

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects