Calling with image_url needs updating

[kirkhofer]

The original code example is incorrect. Calling with an image_url requires a bit of different syntax:

Here is an example:

{
    "messages":[
        {
        "role": "user",
        "content": [
            {"type": "text", "text": "What's in this image?"},
            {
            "type": "image_url",
            "image_url": {
                "url": "https://learn.microsoft.com/azure/ai-services/computer-vision/media/quickstarts/presentation.png"
            }
            }
        ]
        }
    ],
    "max_tokens":300
}

You can see the logic in the OpenAI documentation here

---

Document Details

⚠ Do not edit this section. It is required for learn.microsoft.com ➟ GitHub issue linking.

• ID: f3231140-6a72-c587-61e7-434ec2a52ac9
• Version Independent ID: 68609d85-dc97-bb55-f962-d3e8f113bb96
• Content: How to use the GPT-4 Turbo with Vision model - Azure OpenAI Service
• Content Source: articles/ai-services/openai/how-to/gpt-with-vision.md
• Service: azure-ai-openai
• GitHub Login: @PatrickFarley
• Microsoft Alias: pafarley

[SaibabaBalapur-MSFT]

@kirkhofer Thanks for your feedback! We will investigate and update as appropriate.

[RamanathanChinnappan-MSFT]

@kirkhofer I've delegated this to @PatrickFarley, a content author, to review and share their valuable insights.

[PatrickFarley]

Thank you, Both input formats work. I'll ask the engineer to confirm our choice to recommend the current format.

[PatrickFarley]

I got the go-ahead to match OpenAI's style. PR here https://github.com/MicrosoftDocs/azure-docs/issues/117996

please-close