Result shapes

Responses have no generic artifact envelope, kind discriminator, event ID, cache marker, or completion marker.

AI Name

{
  "name": "Terminal Launch",
  "ticker": "TERM"
}

AI Name returns one generated result. Translation is not part of the response; use the independent Past Coins route when you want related search results.

Past Coins

{
  "requestId": "req_past_coins_01",
  "query": "Terminal Launch",
  "providers": ["solana"],
  "results": [
    {
      "identifier": "solana_tracker:sol:7NfY...pump",
      "name": "Terminal Launch",
      "ticker": "TERM",
      "image": "https://assets.example.com/tokens/term.png",
      "twitter": "https://x.com/terminallaunch",
      "website": "https://terminal.example.com",
      "total_fees_native": 12.4,
      "fee_currency": "SOL",
      "total_fees_usd": 1847.16
    }
  ]
}

query is the normalized term searched by this request. results may be empty. Every coin includes name, ticker, and image; identifiers, platform, social links, provider metadata, and native/USD fee data are included when available. Each request searches one term. If a name search is empty, send a separate request using its ticker rather than adding fallbackQuery.

Crops

{
  "requestId": "req_crops_01",
  "results": [
    {
      "imageIndex": 0,
      "crops": [
        {
          "source": "merged",
          "label": "Terminal announcement",
          "score": 0.94,
          "x": 96,
          "y": 58,
          "width": 548,
          "height": 310
        }
      ]
    }
  ]
}

imageIndex is the zero-based position in the request’s images array. Coordinates and dimensions are source-image pixels. source is prompt_box, ocr, or merged.

OCR

{
  "requestId": "req_ocr_01",
  "results": [
    {
      "imageIndex": 0,
      "text": "TERMINAL LAUNCH\nLIVE TODAY",
      "lines": [
        {
          "text": "TERMINAL LAUNCH",
          "confidence": 0.97,
          "points": [
            { "x": 118, "y": 84 },
            { "x": 612, "y": 84 },
            { "x": 612, "y": 156 },
            { "x": 118, "y": 156 }
          ]
        }
      ]
    }
  ]
}

text is the combined recognized text. Each line includes confidence and source-pixel points. imageIndex maps the result back to the request.

Image Edit

{
  "requestId": "req_image_edit_01",
  "images": ["data:image/jpeg;base64,/9j/4AAQSkZJRgABAQAAAQABAAD..."]
}

images always holds exactly one edited image as a 512x512 JPEG data URL. PNG, JPEG, and WebP inputs with another aspect ratio are scaled to fit and centered on a white canvas, so the output may include white margins. Other image formats are center-cropped to a square. Decode the base64 data after the first comma to get the JPEG bytes.

Get started

API reference

Reference

AI Name

Past Coins

Crops

OCR

Image Edit

​AI Name

​Past Coins

​Crops

​OCR

​Image Edit

AI Name

Past Coins

Crops

OCR

Image Edit