Unsloth is a local AI platform for running, training, and deploying large language models and diffusion models on personal hardware. It supports a wide range of models including DeepSeek, Qwen, Gemma, and FLUX, with fine-tuning that is up to 2x faster and uses 70% less VRAM than standard methods. Unsloth offers a desktop app, a web UI called Unsloth Studio, and a code-based core library, all compatible with NVIDIA, AMD, Intel, and Apple Silicon hardware. It includes features such as LoRA and QLoRA fine-tuning, reinforcement learning, RAG, an OpenAI-compatible API, and agent integrations with tools like Claude Code and OpenAI Codex.

Unsloth
Run and fine-tune LLMs and diffusion models locally.

github.comOpen Source
Upvotes
1
Pricing
Open Source
Category
Generative Code
In the database since
Aug 2026
Overview
About Unsloth
Newsletter
Tools like this change fast
New Generative Code tools launch every week — the newsletter keeps you ahead of what's worth trying.
AI news twice a week
Join 250,000+ readers getting the most important AI news and coolest tools every Wednesday and Friday.