Lemonade v10.8: auto memory management, cloud offload, Omni improvements, and call your local models as MCP tools

Lemonade v10.8 introduces significant enhancements including dynamic VRAM management for automatic unloading of idle models and context sizing based on available memory, improving memory efficiency. A new provider-agnostic cloud offload feature allows integration with OpenAI-compatible services alongside local models, facilitating larger model usage without defaulting to the cloud. Additionally, the MCP gateway enables local models to function as tools for various tasks, expanding usability and integration options for AI practitioners.