Sage docs | Levanto

Works on /decide and on each batch group.

Request

{
  "content": {
    "kind": "image",
    "media": "data:image/png;base64,iVBORw0KGgo...",
    "text": "Ticket: the checkout button is missing on mobile."
  },
  "question": {
    "id": "shows_bug",
    "kind": "yesno",
    "instructions": "Does the screenshot show the reported problem?"
  }
}

Field	Description
`kind`	`"image"`
`media`	Base64 `data:` URI: PNG, JPEG or WebP
`text`	Optional context, such as the ticket for a screenshot

Response

Same shape as text. meta.usage adds image_tokens.

{
  "id": "shows_bug",
  "kind": "yesno",
  "result": { "answer": "yes", "probability": 0.98 },
  "meta": {
    "model": "levanto-sage-2",
    "latency_ms": 251.6,
    "usage": { "input_tokens": 81, "output_tokens": 0, "image_tokens": 313 }
  }
}

On images, Tags always return true or false. Size and option limits: see Limits and errors. Image tokens are charged as input tokens; see Pricing.