Is there a plan to support an image upload feature in Retool AI chat component, especially for RAG use-cases?

We're interested in developing a chatbot that allows users to upload images, enabling the chatbot to generate responses based on those images. Is this feature in the pipeline?

1 Like

Hello @Yashodhar_Meduri,

We currently have a wide variety of options for AI actions related to users uploading images.

We also have several components for building apps that allow for users to upload image files, which we then can then use Retool to encode in base64 and send to the LLM model of choice.

CleanShot 2024-10-17 at 10.33.30@2x

There are also options for setting up chatbots that are able to respond to messages that users enter in. I would recommend watching this video created by our team giving an example of setting up AI text responses in a workflow.

Combining both text inputs and image file uploads is not an 'out of the box' feature that we currently have, but it can be made using the various functional pieces currently in Retool.

1 Like

Hi, just wanted to give an update: as of today this feature request is still in our backlog. The workaround that Jack suggests is currently the only way to do it. I'll respond when I hear any further updates on this!

1 Like