Unsloth

Unsloth is a cross-platform desktop app (Windows, macOS, Linux) that lets you download, run, and fine-tune open-source large language models on your own…

  • AI
Unsloth is a cross-platform desktop app (Windows, macOS, Linux) that lets you download, run, and fine-tune open-source large language models on your own hardware, bundling a notebook-based Studio, a curated model catalog with one-click installs, and accelerated LoRA and QLoRA training so AI developers can work locally without a cloud GPU.

Platform availability

Available on 3 of 10 platforms

  • Unraid (not listed)
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox
  • Helm (not listed)
  • Mac
  • Windows (not listed)
  • Linux
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

86/100 Excellent

Maintained
100last commit 8 days ago
Popular
9879,440 GitHub stars
Easy to install
453 install platforms, install notes
Light to run
No data

Based on 3 of 4 factors.

Availability: platform listings confirmed today

GitHub stars
79,440
Open issues
499
Last commit
2026-11-13
Latest release
v0.1.917-beta
License
Apache-2.0
Activity
steady

Checked 8 days ago - source: GitHub (unslothai/unsloth)

Resources & Compatibility

Requirements not published yet.

First-install notes

  • Set up Docker GPU access first If you are using Docker with an NVIDIA GPU, run the one-time NVIDIA Container Toolkit setup script so the container can use your graphics card. Without it, Docker may not see the GPU and model work can be much slower or fail. source
  • Open ports and back up stored data Use the shown docker run command; it publishes ports 8000 and 8888 for Studio access and local services, and keeps your data in unsloth-workspace and unsloth-studio, which are Docker volumes (saved folders managed by Docker). If you delete those volumes or do not keep the ports open, you may not reach the app and you can lose projects and settings. source
  • Add a password if others can reach it If other devices will open the Studio web page, set UNSLOTH_STUDIO_PASSWORD, a setting passed before the program starts, to a strong password and start with --secure. This required login keeps other computers on your network from seeing or changing your settings. source

Fetched 8 days ago - each note links to its source

Alternatives (2)

  • LLaMA Factory(Nvidia GPU) - LLaMA Factory is an open-source, self-hosted framework for fine-tuning and serving large language models on NVIDIA GPU hardware.
  • Mistral Forge - Transform institutional knowledge into frontier-grade LLMs—without infrastructure burden or cloud lock-in.

Checked 8 days ago - source: AlternativeTo

Common questions

Can I run Unsloth with Docker?

Yes, Unsloth provides an unsloth/unsloth Docker image for Docker-based installation.

Which accelerators does Unsloth support?

It supports multi-GPU setups with NVIDIA, AMD, and Intel GPUs, plus CPUs and the Vulkan backend.

Can I use the local model from another device?

You can access the local models from any device on the LAN, or remotely through secure Cloudflare HTTPS.

Does Unsloth provide an OpenAI-compatible endpoint?

Yes, you can serve Unsloth models through an OpenAI-compatible API.

Can I limit file access when a model executes commands?

Unsloth includes sandboxing to restrict which files the model can access while running commands.

Answers sourced from github.com

Community

Collected 8 days ago - public community threads