In-App Search: You do not need to download model files through a web browser manually. LM Studio features a built-in search bar connected directly to Hugging Face.
lock
Multi-Model Playground & Split View
Prompt Testing: You can send a single prompt to all loaded models simultaneously to compare their response speed, reasoning ability, and accuracy side by side.
extension
Hardware Acceleration & Memory Management
GPU Offloading: If you have an NVIDIA, AMD, or Apple Silicon graphics card, you can slide a selector to push parts of the AI model onto your GPU. This dramatically increases text generation speed.