Gemini 2.0 Flash Error: Basic LLM Chain Node Not Supporting Base64 Image Input

Based on the documentation and behavior you’re describing, it appears that the error “This model does not support images” is coming from the node itself when using the Gemini 2.0 Flash model in the Basic LLM chain node. Even though you correctly pass the image as a base64-encoded string (using the prefix "data:image/jpeg;base64,"), the model you selected isn’t set up to accept image input when called in this way.

What’s Happening

• Basic LLM Chain Node Assumptions:
  The Basic LLM chain node is designed to work primarily with text inputs. When you supply an image (even as base64) in a field that is primarily checked for text, the node verifies whether the selected model supports images.

• Model-Specific Image Support:
  As noted in the documentation, not all models or endpoints in the Gemini API accept image inputs. Although the HTTP request you tried works fine (because you are directly calling the API endpoint that handles image uploads), the Basic LLM chain node is likely using a different integration method that does not support images for Gemini 2.0 Flash.

Recommended Steps

1. Verify Model Compatibility:
   Double-check that the model you’re using is one which supports image inputs. Review the models documentation on the Gemini API for details on which models can analyze images. For example, while Gemini 1.5 Flash models have specific image capabilities, Gemini 2.0 Flash might be limited (or its integration via the node may be limited) to text-only functions.

2. Use the Correct Node/Endpoint:
   If you need to analyze images, consider using the HTTP Request node (as you already tested) or another node specifically designed for image inputs. There may be a dedicated image analysis node or configuration that correctly handles base64 images as required by the Gemini API.

3. Confirm Input Configuration:
   Ensure the image data is passed in exactly as the API expects. Although you are already converting to base64, confirm if there are any additional parameters or different fields required when dealing with images compared to text inputs.

Next Steps

• If you intend to perform image analysis with Gemini 2.0 Flash, you may need to either use the API directly (via the HTTP Request node) or check if a newer version of the integration (or additional configuration) is available for the Basic LLM chain node to handle image inputs.
• If text inputs work fine with the node, then the error message is an indication that the integration does a model check and found that the selected Gemini 2.0 Flash model is not image-enabled when used through that node.

By ensuring you’re using a model endpoint that supports images—and possibly opting for direct API calls when needed—you should be able to bypass the “model does not support images” error.