[Question] VS 2026 Insiders and Foundry local

Hey,

using VS2026 Insiders for quite some time, and lately found that "new" option to manage used models.
Its not about VS itself, more about if I skipped something.

It says "Foundry Local" as model provider, but then I need an API key, instead of a Url, which would probably would make more sense, cause the url seems to change on every start of the server.

So what do I have to fill in there to make it work ? And no, an empty API key does not work.

Start by reproducing the model-provider setup described for VS 2026 Insiders with Foundry Local, including the API-key field and changing server URL. Check the relevant provider configuration guidance or entry point to determine what value is expected; done means documenting the required configuration or confirming that the integration needs a product change.

Written by the indexing model from the issue text.