# Tambaqui

Tambaqui is a self-hosted, OpenAI-compatible AI coding assistant that downloads and runs open-source models (Qwen, DeepSeek, CodeLlama) from Hugging Face locally. It is open source and self-hosted, installable on Unraid.

Canonical page: https://ourlocalhost.com/app/3370/tambaqui

## Description

Tambaqui is a self-hosted, OpenAI-compatible AI coding assistant that downloads and runs open-source models (Qwen, DeepSeek, CodeLlama) from Hugging Face locally. It provides a web chat with web search, streaming responses, and an admin panel, and works with Open WebUI, Continue.dev, LibreChat, and any OpenAI API client, giving homelab users and developers a private LLM backend for coding tasks.

## Platform availability

Available on 1 of 10 platforms: Unraid.

## Health score

36/100 (Low), based on 3 of 4 factors.
- Maintained: 60 (last commit 149 days ago)
- Popular: 0 (0 GitHub stars)
- Easy to install: 35 (1 install platform, install notes, Docker image)

## Links

- Unraid: https://ca.unraid.net/apps/tambaqui-0zgyj9i1dyvgyy?q=Tambaqui
- GitHub: https://github.com/koalitos/tambaqui

## Common questions

### What are the default Docker login credentials for Tambaqui?

The default Docker login is admin for the username and tambaqui for the password.

### Where are the web chat, admin panel, and API served?

Tambaqui serves the web chat and OpenAI-compatible API on port 8000. The admin panel is also on port 8000 under the admin path.

### How can I manage users from the command line?

Use the built-in CLI to create, list, change passwords for, and delete users with commands such as user criar, user listar, user senha, and user deletar.

### How do I get models running in Tambaqui?

You can download models through the web admin panel or select models from the catalog.

---
Source: Our Local Host (https://ourlocalhost.com/app/3370/tambaqui), the self-hosted and open-source app directory.
