โžก๏ธImage to Text Block

Image-to-text spells generate text based on an image.

Title

  • Click the pencil icon next to the blockโ€™s title to edit it.

    • Tip: This is also the blockโ€™s variable name, which can be used to dynamically populate data in other blocks.

Settings

Model

Specify which one of the supported image-to-text models youโ€™d like to use. Currently supported models are:

  • Blip

  • Blip2

  • MiniGPT4

Image

  • Enter the URL of the image or enter a prompt that will create an image.

Task

  • Specify a task for the AI to complete: Caption or Answer Question.

Question

  • If you selected Answer Question, include the question within the text field.

Last updated