telegram-search

Telegram Search provides cutting-edge local search capabilities for your Telegram messages, including vector embeddings, multi-language tokenization, and RAG-enabled Q&A. Run the server with the provided Docker image to expose an HTTP/WebSocket API for indexing, searching, and exporting chat data. Use the integrated bot features to search and export messages via a Telegram Bot, or access the HTTP/WebSocket endpoints for programmatic search and retrieval. The product supports media embedding, image search, and smart summaries of unread messages, making it easy to back up and locate information across languages. Authenticate and configure through the app’s API settings, including embedding and LLM options per account.

Prerequisites:\n- Docker (and optionally Docker Compose) installed on your host\n- Basic familiarity with environment variable configuration for your deployment.\n\nStep-by-step:\n1) Install Docker if not present on your system.\n2) Create a dedicated directory for the Telegram Search deployment.\n3) Run the Docker-based deployment using the command below (adjust env vars as needed):\n\nbash\ndocker run -d --name telegram-search \ -p 3333:3333 \ -e TELEGRAM_API_ID=YOUR_API_ID \ -e TELEGRAM_API_HASH=YOUR_API_HASH \ -e DATABASE_TYPE=postgres \ -e DATABASE_URL=postgresql://postgres:password@localhost:5432/postgres \ -e TELEGRAM_BOT_TOKEN=YOUR_BOT_TOKEN \ ghcr.io/groupultra/telegram-search:latest\n\n\n4) If using Docker Compose, prepare a docker-compose.yml as provided by the project and start services with:\n\n```bash\ndocker compose -f docker/docker-compose.yml up -d\n````\n4) Open the UI at http://localhost:3333 (or the port you configured) to begin configuring API access, embedding, and search options.\n\nOptional: If you need MinIO or other services, follow the repository’s Docker Compose setup to run all components together.\n

Tips and common issues:\n- Ensure your Telegram API credentials and Bot Token are valid; invalid credentials will prevent indexing.\n- If you change environment variables, restart the Docker container to apply changes.\n- The image may require network access to Telegram services; ensure outbound access is allowed.\n- For production, consider mounting persistent volumes for data and media storage instead of relying on ephemeral containers.\n- If you are using a reverse proxy (nginx/traefik), set BACKEND_URL and ports accordingly to expose /api and /ws endpoints.\n- Check the container logs (docker logs telegram-search) for troubleshooting startup issues.\n