Repolode
sdcb

sdcb/chats

A self-hostable frontend and AI gateway for large language models, connecting 22+ model providers through one interface.

Open-source alternative to:ChatGPT

Stars
618
Forks
109
Open issues
10
Last commit
26 days ago

What is it?

Sdcb Chats is a web frontend for large language models that lets you manage many model providers from one place. It supports 22+ mainstream AI providers and exposes Chat Completions and Messages compatible APIs, so existing clients such as Claude Code can route through it.

The project can be deployed quickly with a single Docker command, and native executables are also provided for several platforms. Storage can run on SQLite, SQL Server or PostgreSQL, with local disk or cloud object storage such as AWS S3, Aliyun OSS or Azure Blob for files. Administration covers user permissions, account balance control, rate limiting, audit logs, Keycloak SSO and SMS login.

Additional capabilities include a Docker-based sandbox for code execution, built-in browser and Excel tools, visual input, image generation, and request tracing for diagnosing inbound and outbound API calls. The backend is written in C# and the frontend in TypeScript with React.

Key features

  • 01Unified access to 22+ large language model providers from one interface.
  • 02Chat Completions and Messages compatible API gateway.
  • 03Docker-based sandbox for code execution with built-in browser and Excel tools.
  • 04Request tracing for inbound and outbound API calls.
  • 05Choice of SQLite, SQL Server or PostgreSQL, and local or cloud object storage.
  • 06User permissions, balance control, rate limiting and audit logs.

Use cases

  • Teams that want one web entry point for several LLM providers.
  • Self-hosting a private chat interface with user accounts and balance limits.
  • Routing Claude Code or other API clients through a single gateway.
  • Debugging model API requests with detailed request traces.

Quick start

  1. 1

    Start the container with Docker

    mkdir -p ./AppData && chmod 755 ./AppData && docker run --restart unless-stopped --name sdcb-chats -e DBType=sqlite -e ConnectionStrings__ChatsDB="Data Source=./AppData/chats.db" -v ./AppData:/app/AppData -v /var/run/docker.sock:/var/run/docker.sock --user 0:0 -p 8080:8080 sdcb/chats:latest
  2. 2

    Open http://localhost:8080 and log in with the default account

Things to know

  • The quick-start command mounts the Docker socket and runs the container as root.
  • The main README is in Chinese; the English version is a separate file.

In collections

Topics

FAQ

What is chats?
A self-hostable frontend and AI gateway for large language models, connecting 22+ model providers through one interface. Sdcb Chats is a web frontend for large language models that lets you manage many model providers from one place. It supports 22+ mainstream AI providers and exposes Chat Completions and Messages compatible APIs, so existing clients such as Claude Code can route through it.
Is chats free to use? What license does it have?
chats is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is chats still maintained?
Actively maintained. The last commit to chats was on Oct 12, 2026.
What can I use chats for?
Teams that want one web entry point for several LLM providers. Self-hosting a private chat interface with user accounts and balance limits. Routing Claude Code or other API clients through a single gateway. Debugging model API requests with detailed request traces.
How do I install chats?
Start the container with Docker: mkdir -p ./AppData && chmod 755 ./AppData && docker run --restart unless-stopped --name sdcb-chats -e DBType=sqlite -e ConnectionStrings__ChatsDB="Data Source=./AppData/chats.db" -v ./AppData:/app/AppData -v /var/run/docker.sock:/var/run/docker.sock --user 0:0 -p 8080:8080 sdcb/chats:latest Open http://localhost:8080 and log in with the default account
Where can I get chats?
The source code and installation instructions are on GitHub: https://github.com/sdcb/chats

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects