v0.1.70-beta
📦 unslothView on GitHub →
✨ 13 features
Summary
Unsloth Desktop is released, offering a local AI model development environment for Windows, macOS, and Linux. It features enhanced tool calling, support for new models, faster inference, no-code training, and broad hardware compatibility.
✨ New Features
- Introduction of Unsloth Desktop, a cross-platform application for running and training AI models locally.
- Up to 50% more accurate tool calling with self-healing calls and sandboxed code execution.
- Support for running models like Muse Glimmer 30B, Kimi K3, Qwen3.8, DeepSeek-V4 Flash 0731, and Gemma 4.
- Video generation with MiniMax-H3 and faster image/video creation with diffusion models (up to 2x inference speed).
- Unlimited private web search, Deep Research with cited reports, and RAG capabilities.
- Model export to NVFP4, GGUF, and other formats.
- Remote access to Unsloth models via Cloudflare HTTPS.
- Support for running models on CPU or multiple GPUs across NVIDIA, AMD, Intel, and Mac hardware.
- No-code model training with reduced time and VRAM usage.
- Local models served through an OpenAI-compatible API.
- Integration with OpenAI and Anthropic cloud model providers.
- Sandboxed Python and Bash execution for models.
- Support for LoRAs, reference images, and ControlNet in diffusion model workflows.