← All projects
Training & fine-tuning

XTuner

A training engine for very large AI models built from many expert parts, aimed at big computing clusters.

New this week

What it does

XTuner is software for training large language models, especially "mixture of experts" (MoE) models, which are built from many specialised sub-networks. Version 1 is designed for training extremely large models.

Its makers say it can train models with hundreds of billions of parameters more simply and efficiently than traditional methods, handles long texts, and is optimised for both NVIDIA GPUs and Huawei Ascend chips.

It supports pre-training, fine-tuning with instructions, reinforcement learning, and training models that understand both images and text.

Who it's for

AI research labs and engineering teams training very large models on clusters of accelerators.

Can I use it without coding?

Developers only. It's a training framework for experts with large computing clusters. New to this? Start here.

Can I run it?

Cost
FreeNothing to pay or sign up for.
Runs on
Server
Computer
Made for servers or data centres
Privacy
Runs on your own device: your data stays with you
API key
Not needed

Try it

Using Claude Code, Cursor or another AI coding assistant? Paste the prompt and it will walk you through the setup.New to this?

Before you install

  • Only download from the official GitHub page. Copies with similar names exist for many popular projects.
  • License (Apache-2.0): Permissive: free to use, change and share, including for business.
  • More on staying safe