ZGI is a powerful enterprise-grade AI development platform focused on visual Agent workflow orchestration, advanced RAG systems, and multi-Agent collaboration.
- Multi-model dialogue support with parallel model comparison
- AI Agent integration via @mentions for personalized interactions
- RAG-enhanced knowledge base conversations
- Voice interaction and speech synthesis
- Image recognition and analysis
- File Q&A for PDF, Word, Excel, and more
- Visual debugging with response time and behavior analysis
- State-of-the-art Retrieval Augmented Generation (RAG)
- Multiple vector database support (FAISS, Milvus, Weaviate, Qdrant)
- Customizable recall rate and search parameters
- Rich API ecosystem with RESTful and GraphQL support
- High-availability distributed architecture
- Granular data access control
- High-performance knowledge management
- No-code visual workflow editor
- Flexible node-based design for parallel/serial task execution
- Domain-specific LLM adaptation
- Real-time monitoring and debugging
- Configurable inter-Agent communication
- Comprehensive RESTful APIs with Webhook support
- Detailed documentation, SDKs, and example code
- Modular architecture for custom extensions
- Microservices compatibility
- Multi-environment deployment options
- API-level access control
- High concurrency support
- OpenAI SDK-compatible interface
- Built-in support for OpenAI, Claude, Gemini, LLaMA, Mistral, Command R+
- Multi-model comparison and switching
- On-premise deployment options
- Real-time token tracking and cost management
- Rate limiting and access control
- Comprehensive logging and auditing
- Hierarchical organization and project management
- Fine-grained permission control
- Data isolation and encryption
- SSO support (OAuth, LDAP, SAML)
- Audit logging and compliance
- Private deployment options
- Backup and recovery mechanisms
✅ Enterprise-Ready: Complete organization, project, and permission management with private deployment options ✅ Smart Knowledge Management: Advanced RAG system with multimodal knowledge base and semantic search ✅ Visual AI Workflows: No-code Agent orchestration with multi-Agent collaboration ✅ Extensive Model Support: 1000+ models via LLMS Gateway with OpenAI SDK compatibility ✅ Powerful API Ecosystem: Standardized APIs with comprehensive documentation ✅ Security-First: Fine-grained access control with end-to-end encryption ✅ High Performance: Cloud-native architecture with distributed storage
- Ensure you have yarn or bun installed.
-
Clone the repository:
git clone https://github.com/zgiai/zgi.git cd zgi -
Install dependencies:
yarn install # or bun install -
Start the API server:
cd api python run.py -
Start the development server:
yarn dev # or bun dev
zgi/
├── api/ # Backend API services
├── frontend/ # Web interface
├── desktop/ # Desktop application
├── sdks/ # Software Development Kits
├── docs/ # Documentation
├── examples/ # Usage examples
└── docker/ # Docker configuration files
- 🎙️ Chat with PDF: Coming soon
- 📚 Knowledge Spaces: Coming soon
- 🌐 Multi-language Support: Expanding language support
- 📱 Mobile App: Native mobile applications
We welcome contributions! Please see our Contributing Guidelines for details.
To protect your privacy, please avoid posting security issues on GitHub. Instead, send your questions to security@zgi.ai and we will provide you with a more detailed answer.
This repository is available under the MIT License. See the LICENSE file for more info.
Special thanks to all contributors and the open-source community for making this project possible.