Generate text prompts from an image

Analyzes an uploaded image and generates multiple text prompt suggestions that could be used to recreate a similar image with the imagine endpoint. The image can be provided as a URL or base64-encoded data. Returns up to four prompt suggestions per image.