Welcome to DeepExtension Documentation¶

DeepExtension is an AI infrastructure platform built to help enterprises easily manage the entire lifecycle of large language model (LLM) development—from data preparation to fine-tuning, evaluation, and deployment.

Our mission is to make domain-specific AI development accessible, collaborative, and scalable, especially for teams with limited AI expertise or compute resources.

Whether you're an AI engineer or a business expert, DeepExtension offers a shared environment to build impactful models using modern techniques like PEFT, GRPO, and human-in-the-loop workflows—all through a modular, web-based interface.

Getting Started¶

Start your journey with DeepExtension by following the Quick Start Guide.

Table of Contents¶

Supported Languages¶

This documentation is available in:

English (Default)
Deutsch
中文

About DeepExtension¶

DeepExtension is purpose-built for enterprise-grade LLM training and deployment. It simplifies complex AI workflows through:

A fully web-based interface
Modular orchestration of training, evaluation, and deployment
Integration of developer-defined training logic
Support for PEFT, SFT, PPO, GRPO, and more
Real-world case templates from biotech, materials science, industrial design, and beyond

It is especially well-suited for:

Organizations with limited GPU infrastructure
Teams that want to start quickly using best-practice templates
Companies looking to bridge the gap between AI developers and business users

Contact Support¶

For help or questions, visit our Support page or reach out to our team.