Set Up a Local GitHub Copilot‑Like AI Assistant in 5 Minutes

Introduction

Tabby is an open‑source, self‑hosted AI coding assistant that enables teams to run a local LLM‑driven code‑completion service.

Deploy with Docker

Ensure Docker, Nvidia drivers, and the Nvidia Toolkit are installed on the host.

Start Tabby with Docker.

docker run -it --gpus all \
  -p 8080:8080 -v $HOME/.tabby:/data \
  registry.tabbyml.com/tabbyml/tabby \
  serve --model StarCoder-1B --chat-model Qwen2.5-Coder-1.5B-Instruct --device cuda

Register an Account

After the container starts, open the URL shown in the startup logs (e.g., http://{ip}:8080) and create an admin account. The home page displays statistics such as the number of generated code snippets and their usage rate.

Configure Visual Studio Code

Install the Tabby VSCode extension.

The extension is available on the Visual Studio Marketplace and Open VSX.

ext install TabbyML.vscode-tabby

After installation, click the Tabby button in the lower‑right corner of VS Code, then paste the token shown on the login page to connect the client.

Code Completion Demo

When editing a Python file, Tabby suggests multi‑line completions and full functions in real time. A quick‑sort implementation is completed instantly, demonstrating low latency and useful suggestions.

References

Qwen2.5‑Coder‑1.5B‑Instruct: https://huggingface.co/Qwen/Qwen2.5-Coder-1.5B-Instruct-GGUF

Tabby VSCode extension (Visual Studio Marketplace): https://marketplace.visualstudio.com/items?itemName=TabbyML.vscode-tabby

Open VSX: https://open-vsx.org/extension/TabbyML/vscode-tabby