OpenChat
AI chat for your entire organisation.
The equivalent of a private ChatGPT deployed on your own infrastructure — multi-modal, RAG-powered, and fully under your control.
Everything your organisation needs
Multi-modal
Interact with AI using text, voice, images, and documents. Full multi-modal support powered by locally hosted models.
Knowledge Management
RAG-powered document chat over your internal knowledge base — connect Google Drive, SharePoint, or custom data sources.
Agentic Capabilities
Tool use via MCP integrations. Connect to Gmail, Teams, CRMs, and custom APIs without data leaving your premises.
Collaboration
Organisation-wide deployment with shared workspaces, prompt libraries, and team-level AI assistants.
Role-Based Access Control
Granular RBAC for every team. Control who can access which models, knowledge bases, and integrations.
API Gateway & Model Routing
One OpenAI-compatible endpoint for every model via LiteLLM. Route requests across local and external backends with per-team keys, rate limits, and usage tracking.
Technology stack
| Component | Role |
|---|---|
| Open WebUI | Chat interface and orchestration |
| vLLM | High-performance LLM inference |
| Whisper | Automatic speech recognition |
| Kokoro | Text-to-speech engine |
| ComfyUI | Image generation engine |
| OpenTerminal | Sandboxed terminal and code execution |
| Docling | Document parsing and conversion |
| LightRAG | Graph-based RAG assistant |
| Qdrant / Neo4j | Vector store & knowledge graph |
Ready to deploy OpenChat?
Book a discovery session to scope your deployment, hardware requirements, and integration needs.
Book a Discovery →