Inference
Images
Use POST /v1/chat/completions with the quickstart setup.
Send an image
Replace messages in your request with this array:
[
{
"role": "user",
"content": [
{"type": "text", "text": "Describe this image."},
{
"type": "image_url",
"image_url": {"url": "data:image/jpeg;base64,<encoded-image>"}
}
]
}
]Replace <encoded-image> with base64 image bytes; use image/jpeg or
image/png to match the file. Add more image_url items for multiple images.
Streaming and usage
Set stream: true to stream the text response. Image tokens count toward
the input context and are charged at the model's input rate.