a fast, self hosted web interface for chatting with large language models
NextChat is an open source web interface for chatting with large language models. It gives a team one private chat front end that talks to the model providers they already pay for: conversations organised in a sidebar, reusable system prompts and a prompt template library, per conversation model and provider switching, and transcript export. It is a front end rather than a model host, so it calls a provider API and runs on a small CPU only machine with no GPU. OpenAI and Azure OpenAI, Anthropic, Google Gemini, DeepSeek, Moonshot, Alibaba Qwen, ByteDance, iFlytek, xAI, ChatGLM and SiliconFlow are supported, along with any OpenAI compatible endpoint.
cloudimg ships the upstream release pinned by image digest, so the deployment is exact and reproducible and never silently upgrades. No provider API key is ever baked in: every credential ships empty and you supply your own, or each person supplies theirs in the browser. Because NextChat switches its own access check off whenever no access code is set, a unique access code is generated on each machine at first boot and the service carries a preflight guard that refuses to start without one, so an instance is never left open on your provider credits. Add a fully patched base with unattended security upgrades, a paired deployment guide, and 24/7 cloudimg support.
Real screenshots taken while testing this image against its deployment guide.