chat

Why we included this project

This project gives a small team a single self-hosted chat interface that can talk to OpenAI and Azure-hosted ChatGPT, Claude, Gemini, and local Ollama models, so users don't have to switch apps when they switch backends. It also handles the operational details that matter in production: the first registered user becomes admin, per-user rate limits keep heavy usage in check, and an OpenAI-compatible gateway issues revocable virtual API keys with streaming proxy, request auditing, and time-limited request/response samples. Conversation snapshots can be exported as shareable static pages (ShareGPT-style) or resumed later, and a full-text searchable snapshot directory makes it easy to find and revisit past sessions. If you need to control access, track usage, and keep model choice flexible without paying per-seat SaaS fees, this is a reasonable starting point.