Ollama Setup Instructions

Download and Install Ollama

Download the current version of Ollama for your platform from the official website: HERE. There’s a standard installer package for each operating system (Windows or macOS).

After the Installation

Once the installation is complete, you need to pull a vision-capable LLM. Most models are multiple gigabytes, so the download will take some time depending on your internet connection speed.

Our recommendations for use with LrGeniusAI via Ollama:

• gemma3:12b-it-q4_K_M or gemma3:4b-it-q4_K_M - The open source version of Google's Gemini models.
• qwen3-vl:4b-instruct-q4_K_M or qwen3-vl:8b-instruct-q4_K_M - Another very strong AI vision model
• llava - Popular vision model
• minicpm-v - Smaller, faster option for basic analysis

You can pull and use any vision-capable model available in Ollama.

Pull an AI Model

You need to download at least one AI model to use Ollama with LrGeniusAI. Recommendations are gemma3 or qwen3-vl. Try multiple models to find the best combinaton of results and performance for your use-case.

See the Ollama model list online

Run at least one of the following commands:

ollama pull gemma3:4b-it-q4_K_M

ollama pull gemma3:12b-it-q4_K_M

ollama pull qwen3-vl:4b-instruct-q4_K_M

ollama pull qwen3-vl:8b-instruct-q4_K_M

ollama pull llava

ollama pull minicpm-v

The download will take some time depending on your internet connection speed.