My nano banana HTTPdoesnt work with my current json structure of text prompt +base 64 of an image

[Blaise_Calinaud]

Describe the problem/error/question

I’m trying to create images in bulk for an e-commerce that needs a specific product variable look and the right logo. No matter how I use nano banana though gemeni API or though replicate API, either it doesnt read the input ofthe logo and necessary model of product variable to follow or it simply blocks like on the error below. My prompt text (in french but it’s not important for the topic) + base64 image transformed 2 nodes behind are all working and well integrated into the json expression. However, I don’t understand what is wrong in this error message ?

Also, if you have recommendations for other ways of executing a workflow of this style (eg. using an agent instead HTTP ) , I’m keen on that too

What is the error message (if any)?

Please share your workflow

Share the output returned by the last node

Information on your n8n setup

• n8n version:
• Database (default: SQLite):
• n8n EXECUTIONS_PROCESS setting (default: own, main):
• Running n8n via (Docker, npm, n8n cloud, desktop app): n8n cloud
• Operating system: windows

[jabbson]

Hey @Blaise_Calinaud hope all is good. Welcome to the community.

Looking at the expected request schema for this request

Nano Banana | Google | Image Editing | schema

It appears that the request you are sending to replicate is wrong. The body/json you have to be sending is as follows:

{
  prompt: <string>,       // A text description of the image you want to generate
  image_input: <array>,   // Input images to transform or use as reference (supports multiple images), Default: []
  aspect_ratio: <string>, // Aspect ratio of the generated image, Default: "match_input_image"
  output_format: <string> // Format of the output image, Default: "jpg"
}

[Blaise_Calinaud]

it works thank very much !

[jabbson]

kindly mark the answer as Solution. thank you.

cheers!