Unsloth Studio: Open-Source No-Code UI for LLM Fine-Tuning with 70% Less VRAM
Unsloth AI released Unsloth Studio on March 17, an open-source, no-code web UI for training, running, and exporting LLMs locally. It runs 100% offline on Mac, Windows, and Linux.
Studio can train 500+ models 2x faster with 70% less VRAM and no accuracy loss, supporting text, vision, TTS audio, and embedding models. It runs GGUF and safetensor models locally, and includes a Data Recipes feature — a visual node-based workflow for transforming unstructured documents (PDF, CSV, DOCX) into structured training datasets using Nvidia's DataDesigner.
The platform includes self-healing tool calling, web search, and code execution capabilities, letting LLMs run code in a sandbox to calculate, analyze data, test code, and verify answers. This makes it practical for building and fine-tuning agent models that need tool-use capabilities.
Unsloth Studio is trending on Hacker News (144 points) and was showcased at Nvidia GTC 2026. For anyone building custom agent models — whether for coding, search, or enterprise workflows — Studio removes the infrastructure barrier entirely.
GitHub: https://github.com/unslothai/unsloth
Website: https://unsloth.ai
← Back to all articles
Studio can train 500+ models 2x faster with 70% less VRAM and no accuracy loss, supporting text, vision, TTS audio, and embedding models. It runs GGUF and safetensor models locally, and includes a Data Recipes feature — a visual node-based workflow for transforming unstructured documents (PDF, CSV, DOCX) into structured training datasets using Nvidia's DataDesigner.
The platform includes self-healing tool calling, web search, and code execution capabilities, letting LLMs run code in a sandbox to calculate, analyze data, test code, and verify answers. This makes it practical for building and fine-tuning agent models that need tool-use capabilities.
Unsloth Studio is trending on Hacker News (144 points) and was showcased at Nvidia GTC 2026. For anyone building custom agent models — whether for coding, search, or enterprise workflows — Studio removes the infrastructure barrier entirely.
GitHub: https://github.com/unslothai/unsloth
Website: https://unsloth.ai