Vision (image input)

Send images alongside text for OCR, screenshots, diagrams and visual questions.

{
  "model": "YOUR_VISION_MODEL",
  "messages": [{ "role": "user", "content": [
    { "type": "text", "text": "Describe this image." },
    { "type": "image_url", "image_url": { "url": "data:image/png;base64,BASE64_IMAGE" } }
  ] }]
}

Check support first

Use the capabilities returned by /v1/models. A model without the vision capability will reject image input.

Image formats

Use a reachable HTTPS URL or a Base64 data URL for local files. Keep payloads within the limits shown by your plan and resize large images before encoding.

Multiple images

Add multiple image parts to the same user content array when the model supports it. Image tokens count toward your normal plan usage.

Need help?

Keep the request ID and contact Torve support without sharing your API key or private data.

Contact support