User reports LM Link web search failures despite local setup

A Reddit user describes configuring LM Studio with Gemma and QWen models alongside a local SearXNG container, successfully linking them via rzk/searxng-search to enable internet data retrieval. While the iPhone interface can interact with the desktop, any actions requiring web searches fail completely.

• The user verified that both models retrieve data from the Internet locally.
• LM Link was enabled to allow iPhone interaction with the home desktop.
• Web search functionality fails entirely within the LM Link environment.

The post questions whether LM Link is too raw and unfinished for reliable use.