DEEIX Chat is an open-source enterprise AI platform in Go with multimodal chat, model routing, files, tools, billing, identity, and operations. It supports local or Docker deployments with PostgreSQL/Redis or SQLite, and a Next.js 16 frontend.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (2 so far).
What it is
DEEIX Chat is an open-source, deployable AI platform for individuals, teams, and enterprises that need long-term, stable, and unified access to multiple model providers. It provides one clear entry point for multiple upstream models and providers, integrating multimodal chat, model routing, files and RAG, MCP tools, usage billing, identity, audit logs, and operational controls into one product.
How it works
The system uses a split frontend/backend design with a single-runtime Go backend and a Next.js 16 frontend. The backend handles APIs, authentication, model routing, files, billing, and audit, while the frontend serves static assets. Data and storage options include PostgreSQL + pgvector or SQLite + sqlite-vec, Redis or in-memory cache, and local filesystem or S3-compatible storage. Optional file processing via built-in extractors (Tika, Docling, OCR) can be added as services.
Getting started
Quick Start
The README links to a Quick Start guide. Local development steps include:
- cp config.example.yaml config.yaml
- pnpm install
- cp frontend/.env.example frontend/.env.local
- pnpm dev
For Docker deployment, there are profiles:
- Lightweight: SQLite with docker-compose.sqlite.yml
- Default: external PostgreSQL and Redis
- Full: PostgreSQL + Redis containers with docker-compose.full.yml
Environment and config notes:
- NEXT_PUBLIC_API_BASE_URL for frontend API requests (default http://127.0.0.1:8080 in local dev)
- APP_ENV can be dev/development or prod/production
- Default app image: ghcr.io/deeix-ai/deeix-chat:latest
Docker deployment specifics
- The default compose starts the app; data paths include /app/data/deeix.db and /app/storage
- Optional services (Tika, Tesseract OCR, Docling) can be started with dedicated docker-compose files
Startup checks
Health endpoint: curl http://localhost:8080/healthz Config check: docker compose exec app ls -l /app/config.yaml
Configuration priority
Environment variables > config.yaml > defaults
Separated deployment
Guidance to set public URLs for frontend and backend when served from different origins.