supportImage property
Inference: whether the model is created with vision. Embedders: the
media the caller declared (getActiveEmbedder(supportImage:)), widened
by core to everything the cached embedder already served.
Implementation
final bool supportImage;