Echo Chat (the swuecho/chat project) is an open-source web application for giving a team shared access to large language models through one self-hosted chat interface. It supports OpenAI and Azure OpenAI, Anthropic Claude, Google Gemini and local Ollama models behind a single UI, with user accounts and rate limiting included. The GitHub repository has roughly 570 stars and is actively maintained, with updates in October 2026.
Sharing model access across a team usually means handing out individual accounts or API keys. Echo Chat instead runs as a small SaaS-style service you host yourself: users register, the first registered user becomes the administrator, and the admin manages members and quotas while everyone chats through the same web front end. Conversations automatically carry recent context, system prompts can be set per conversation, and usage is rate-limited per user (the default cap is configurable) so one heavy user cannot drain the shared API budget. Because it speaks to several providers — including self-hosted Ollama models — a team can mix cloud and local models behind one login. It is a pragmatic way for small organisations, labs and clubs to pool LLM access under their own control.
Deployment is a single self-hosted service backed by a database for users and chat records, and screenshots in the repository show the interface handling long conversations and code across models. For administrators, per-user call quotas and free choice of provider keep spending predictable while still letting each member pick the model that fits the task at hand.
Echo Chat is free and open source under the MIT license. You pay only for your hosting and whatever model API usage your team generates.
Small teams, research groups and communities that want a shared, self-hosted chat front end for several LLM providers, with basic user management and cost controls built in.