Tabby is a self-hosted, open-source AI coding assistant designed as an on-premises alternative to hosted coding tools. It operates without requiring a DBMS or cloud service and is built to run on consumer-grade GPUs.
Project overview
It offers a self-contained deployment path for code completion and developer chat that does not require a DBMS or cloud service.
Project type
AI Coding
Use cases
Coding & Development
Deployment
Docker / Docker Compose
License
License pending
Best for
Development teams seeking an on-premises AI coding assistant that supports consumer-grade GPUs and requires no DBMS or cloud service.
Key capabilities
Exposes an OpenAPI interface to integrate with existing infrastructure such as cloud IDEs.
Utilizes repository-level context to provide relevant code suggestions.
Limitations and risks
Operation requires loading and configuring specific LLMs, such as StarCoder, to serve requests.
Using GPU acceleration requires appropriate hardware, such as consumer GPUs with CUDA support.
Getting started
Setup is categorized as easy due to an available 1-minute Docker quickstart guide.