LM Studio Setup Instructions

Download and Install LM Studio

Download and run the installer: Here

Setup LM Studio for use with LrGeniusAI

Start LM Studio
Go to the Developer Tab
Set status to: running
In the Server Settings: Activate Just-In-Time Model Loading

Pull an AI model

Go to the Discover Tab: Search and download the AI model(s) you want to use. Recommended are:

google/gemma3-4b - Google's open source model
google/gemma3-12b - Bigger version of Google's open source model‚
google/gemma3n-e4b - Reduced size (faster) version of gemma3:4b (macOS only, at the moment)
qwen/qwen3-vl-4b - The current vision model from Alibaba
qwen/qwen3-vl-8b - Bigger size of qwen3-vl-4b

There are different sizes for each model (4b, 8b, 12b). For optimal performance and results: Use the biggest model which offers "Full GPU offload possible". Bigger means more intelligent, but requires more hardware resources.

macOS

If you are running LM Studio on macOS you should models with MLX format. Those models are optimized for Apple Silicon and offer better performance.