A model is missing from the picker, and what its labels mean

The picker leaves out any model its provider has confirmed cannot use tools, and filters by provider, by data policy and by what is usable; a model the provider said nothing about is still listed.

Last checked against the app on · View as Markdown

Steps

  1. Open the model button in Chat and clear the search box and any filter chips you set: "Sees images", "Recommended", "Favourites" or a data-policy setting.
  2. Check the provider has a key or is running. Cloud models appear once a key is saved for their provider, and local models once Ollama or your local server answers.
  3. Know the one permanent exclusion. A model its provider has confirmed cannot call tools is never shown, because NeuralVault drives chat through a tool loop. A model whose provider simply did not say is still shown, which is why local models are listed.
  4. Read the labels on a row. The data-handling label says "no training or logging", "logs data" or "trains on data", and is left off when the vendor’s policy is not known. "Sees images" and the capability marks come from what the provider published. A missing mark means the provider published nothing, which is not the same as no.
  5. Read the AI score only where it is shown. It is part of NeuralVault routing for a paid account, so a model you reach on your own key has no score.

How to tell it worked

The model you wanted appears, or you know which of these explains why it cannot.

If it does not work

The list is empty or says it could not be read.
The button’s hover text gives the reason. If the engine banner is showing, deal with that first.
A model is listed but cannot answer.
The row says why. Choose another, or see the guide on chat not answering.
A local model you pulled is not listed.
See the guide on local model problems.