Automatic video indexing
A background service watches for new video files and sends them into the analysis pipeline, so newly added media can be processed without manual reindexing steps.
Edit Mind
Edit Mind is a local-first video knowledge base that indexes video libraries with transcription, frame analysis, and semantic embeddings, then lets users search content in natural language. It is self-hosted with Docker Compose and described as privacy-preserving, though the project is still in active development and not yet production-ready.
Overview
Edit Mind is a local-first video knowledge base for indexing personal or self-hosted video libraries. It analyzes video files with transcription, frame analysis, and multi-model embeddings, then lets users search the library in natural language.
The repository describes the project as running fully locally for privacy, with Docker Compose used to start the services. It is currently in active development and not yet production-ready, so the source explicitly notes that users should expect incomplete features and occasional bugs.
Core features
Multi-modal analysis
The analysis pipeline extracts transcription, frame-level signals, face recognition, object and text detection, scene analysis, and multi-model embeddings to build richer metadata for each video.
Semantic scene search
Search works with natural-language queries and can target whole videos or specific scenes, making it easier to find moments by meaning rather than filename.
Vector search with ChromaDB
The system uses ChromaDB for vector-based retrieval, supporting semantic matching across the indexed video library.
Docker-first setup
The project is built for containerized deployment with Docker Compose, and the README provides a separate CUDA compose file for NVIDIA GPU setups.
Where it fits
Private personal video archive
A solo user can point Edit Mind at a video folder and build a searchable archive without sending media to a hosted service.
Find specific moments in long recordings
People managing a large library can search for moments by spoken words, faces, objects, or scene content instead of remembering filenames or timestamps.
Self-hosted media search
Teams or hobbyists who self-host infrastructure can run the system in Docker on a server and keep the workflow inside their own environment.
Review extracted video metadata
Users who want to inspect how the pipeline interprets media can rely on the indexed metadata, including transcription and visual analysis, to review what was extracted.
Pros and Cons
Pros
- Runs fully locally, which the README positions as a privacy-preserving approach.
- Supports natural-language search across videos and specific scenes.
- Combines transcription, face recognition, object and text detection, and scene analysis into one indexing flow.
- Works with Docker Compose and includes guidance for NVIDIA GPU users.
- The self-hosted version is described as free.
Cons
- The project is still in active development and not yet production-ready.
- Setup depends on Docker and a configured media folder, so it is not a zero-config app.
FAQ
How is Edit Mind deployed?
Edit Mind is designed to run locally with Docker Compose. The README says it works on any computer or server with Docker installed, and the setup guide uses Docker Desktop plus environment variables for configuration.
What does Edit Mind do with my video library?
The README describes indexing videos with transcription, frame analysis, face recognition, object and text detection, scene analysis, and multi-model embeddings. It then supports natural-language search over videos or specific scenes using ChromaDB.
Is Edit Mind free?
The repository README says the self-hosted version is free. It also mentions a separate commercial desktop app for macOS and Windows with one-click installation, but the source shown here does not list pricing details for that app.
What do I need to get started?
The setup section says you need Docker Desktop, a media folder shared with Docker, and a choice of model setup such as Ollama or Gemini API. It also notes that NVIDIA GPU users can use docker-compose.cuda.yml.
Is Edit Mind production-ready?
The README says the project is in active development and not yet production-ready, with incomplete features and occasional bugs.
Quick Facts
- Category
- Video knowledge base
- Deployment
- Self-hosted with Docker Compose
- Primary workflow
- Index videos, then search them in natural language
- Core technologies
- ChromaDB, PostgreSQL, React Router V7, Node.js, Python, Whisper
- Platform fit
- Any computer or server that can run Docker
- Project status
- Active development; not production-ready
Edit Mind 替代品
dumppp
dumppp is a mobile capture app for Notion that focuses on rapid note entry, context switching, offline capture, and lightweight property updates. It helps Notion users record thoughts quickly on phone without the usual navigation and loading friction.
UStack
UStack 是一个多语言目录，用于发现和比较 AI 工具、SaaS 产品、开发者工具和创意软件。收录 4,000+ 条目，覆盖 150+ 类别，并每日更新。
Agentset
Agentset 是一个开源平台，用于在私有或内部知识库之上构建 AI 聊天和搜索体验。支持带引用的生产级 RAG、图像/图表/表格等多模态文档，并提供免费版、Pro 版和企业版方案。
Firecrawl
Firecrawl 是面向 AI 代理和开发者的 web 上下文 API，可将实时网站转换为干净的 Markdown 或结构化数据。支持搜索、抓取、爬取、监控和浏览器交互，适用于研究与自动化工作流。
Afterpage
Afterpage is a document organizer for iPhone and iPad that helps you scan, import, search, and file paperwork. It uses on-device text recognition and pattern-based suggestions to make archived documents easier to find later.
MealJar
MealJar is a meal planning and recipe organization app for families. It helps users plan weekly meals, store recipes from multiple sources, and coordinate grocery lists with household members.