Changing to a vision model doesn't automatically allow for images with automators

When we go to Anthropic Sonnet 3.5, it doesn't seem to allow for images

I can't get any of the fireworks AI models working either with images

I posted the same issue in the fireworks AI provider thread, and the Amazon Bedrock Provider, and not sure if those issues are related. I cannot use any stable diffusion images on Fireworks AI, and I cannot use any images at all on Amazon Titan.

I have not yet tried Anthropic as a provider, but I can if that will help with testing.