Unsloth Studio Launches in Beta, Letting Users Train 500+ AI Models Locally at Half the Speed With 70% Less VRAM

Unsloth Studio launches in beta, enabling users to train over 500 open-source AI models locally on Windows, Linux, and macOS up to 2x faster while using 70% less VRAM, with features including reinforcement learning, multi-GPU support, and API inference compatibility with tools like Claude Code.

• Unsloth Studio is a web UI now available in beta that allows users to locally run and train open-source AI models including Gemma 4, Qwen3.6, DeepSeek, and gpt-oss on Windows, Linux, and macOS.
• The platform supports training over 500 models up to 2x faster with up to 70% less VRAM, and offers features like reinforcement learning, tool calling, code execution, multi-GPU support, and an API inference endpoint compatible with tools like Claude Code.
• Unsloth Studio is installable via a single command on macOS, Linux, WSL, and Windows, and operates under a dual-license model combining Apache 2.0 for the core package and AGPL-3.0 for the Studio UI, with 64.7k GitHub stars and 216 contributors.