GitHub radar
Unsloth Studio: Train and Tune Large Language Models Locally!
Run & tune language, diffusion, TTS & embedding models locally — web UI, drag-and-drop dataset builder, no cloud needed.
“Unsloth: The first desktop app to run and train models.” Unsloth Studio (run with unsloth studio) provides a web UI for browsing models, downloading and running LLMs, diffusion, TTS, and embedding models locally. Datasets can be built via a drag-and-drop UI, with support for PDF, CSV, DOCX file types for model fine-tuning without code. Available for Windows, macOS, and Linux. Supports NVIDIA, AMD, Intel GPUs, CPUs, Vulkan. Optional integration with ChatGPT / other cloud services. Benchmarks of model performance for Google (Gemma), OpenAI (gpt-oss), Alibaba (Qwen), and Meta (Llama). Fine-tuning 2x faster, 70% less VRAM than traditional pipelines. MoE model preparation 12x faster, 35% less VRAM. 75 000+ stars on GitHub.
Why a vibe-coder should care
Want to use a language model locally, but don’t have a cloud subscription? Use Unsloth Studio to get a functional local UI. Want to fine-tune a model on your company docs, customer Q&A, writing samples? Do it yourself with a GUI and drag-and-drop dataset builder for PDFs and spreadsheets, without code.
How to install
Copy this and send it to your agent — Claude Code, Codex, any of them:
Install Unsloth from https://github.com/unslothai/unsloth — follow the README, run `unsloth studio`, open the web UI in the browser, and show me the list of available models. Ask if you need permissions or additional steps.
Small models run on a regular laptop, no GPU required. For large models and fine-tuning, a GPU is needed: NVIDIA, AMD, Intel, or an Apple Silicon Mac.
Open on GitHub▌ More finds