How to Send an Image to the GPT-4o Module for Recognition

[shimon9]

Hi all,

I'm working on a Blazor Web App (version 9.0) using OpenAI.Client 2.2.0-beta.4, and I’d like to send an image to the GPT-4o model to identify and describe its contents.

Currently, I'm sending the image as a base64 string, but I receive the following response:

"I'm unable to analyze images directly. If you describe the image to me, I can help analyze the contents based on your description."

Is there a correct way to send images to GPT-4o for recognition? Should I be using a different method or encoding format to enable image analysis?

Any guidance would be greatly appreciated.

[shimon9]

Just an update, here is the solution.
https://platform.openai.com/docs/api-reference/responses/create?lang=csharp