Skip to main content
Mix text and image_url inside a message’s content to let the model understand an image (image input, not image generation). 
curl https://tokendog.io/v1/chat/completions \
  -H "Authorization: Bearer $TOKENDOG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model":"gpt-5",
    "messages":[{"role":"user","content":[
      {"type":"text","text":"Describe this image."},
      {"type":"image_url","image_url":{"url":"https://example.com/cat.png"}}
    ]}]
  }'
Vision-capable models (e.g. gpt-5, gemini-2.5-flash, claude-sonnet-4-6) accept image input.