An enterprise AI workspace for teams, DEEIX Chat unifies multimodal chat, model routing, file/RAG workflows, and operational controls. It simplifies working with multiple AI models and providers, offering a centralized platform for efficient AI integration and management.

How It Works

Built with a Next.js frontend and a Go backend, DEEIX Chat prioritizes simple deployment and efficient static delivery. Its architecture features a robust Model Control Plane for dynamic routing, priority-based model selection, and comprehensive provider protocol support (OpenAI, Gemini, xAI, etc.). The system integrates advanced file handling, RAG capabilities with OCR, and a layered backend structure for maintainability.

Quick Start & Requirements

• Local Development: Clone the repository, configure config.yaml, and run backend (make run) and frontend (pnpm dev) separately. Requires Go 1.26, Node.js/pnpm, Docker.
• Docker Deployment: Utilize docker-compose.full.yml for a self-contained environment including PostgreSQL and Redis. Configuration can be managed via config.yaml or environment variables.
• Documentation: Swagger UI available at http://localhost:8080/swagger/index.html.

Highlighted Details

• Unified Workspace: Supports multimodal chat, file management, RAG, and tool integration across diverse AI models.
• Advanced Model Routing: Dynamic priority and weight-based routing, capability mapping, and circuit breakers.
• Extensive Provider Support: Integrates with OpenAI, Anthropic, Google/Gemini, xAI, and custom OpenAI-compatible endpoints.
• Comprehensive Features: Includes media generation, request governance, robust memory/context management, and detailed billing/payment systems (Stripe integration).
• Enterprise-Grade Security: Features SSO/OIDC/OAuth, 2FA/TOTP, trusted devices, and encrypted secrets.

Maintenance & Community

• Contact: support@deeix.com, Telegram: t.me/deeix_chat.
• Resources: Links to CONTRIBUTING.md and SECURITY.md are provided within the repository.

Licensing & Compatibility

• License: Apache License 2.0.
• Compatibility: Permissive license suitable for commercial use and integration into closed-source applications.

Limitations & Caveats

The README does not explicitly detail known bugs or alpha status. Optional file processing services (Tika, Tesseract, etc.) require separate Docker setup. Production deployments necessitate careful configuration of security settings, including strong secrets and HTTPS.