Actions21
- Token Actions
- Recognition Actions
Recognition → PopularCaptcha Classification
AI-generatedSummary
Sends PopularCaptcha image challenge data (single image and/or up to nine grid tiles) to the CaptchaSonic API for classification and returns the API's inline result in a single call, with no polling.
Inputs
- type (required) — Task Type; the only option is PopularCaptcha Classification (sent to the API as
PopularCaptchaImage). - websiteURL — URL of the page hosting the CAPTCHA. Only sent if non-empty.
- websiteKey — Sitekey of the CAPTCHA widget. Sent as both
websiteKeyandwebsiteKEYif non-empty. - image — Base64 image of the CAPTCHA. Added first to the
queriesarray. - imageBackground — Base64 background image, for slide-type challenges.
- question — Instruction text for the challenge, e.g. "Select all images with cars".
- body — Raw body. If it parses as a JSON object, its fields are merged into the task payload (e.g. extra API fields), overriding earlier fields of the same name. Otherwise it is sent as a raw
bodystring. - images 1 … images 9 — Nine optional Base64 grid tile images. Empty ones are skipped; non-empty ones are appended to
queriesin order.
Output shape
One output item per input item, shaped { success: true, data: <CaptchaSonic API createTask response> }, paired to the input item.
The exact fields in
dataare determined by the CaptchaSonic API and not defined in the node. API errors (errorId non-zero) throw an error. With Continue On Fail enabled, the item is instead{ error: message }. Requires CaptchaSonic API credentials (API key).
Examples
Example 1: Classify a 3x3 grid challenge captured by an earlier node
Resource Recognition, Operation PopularCaptcha Classification; set Question to the prompt text and map tile Base64 strings into Image 1–9.
Example 2: Pass extra API-specific fields
Put a JSON object such as {"questionType":"..."} in Body to merge it into the task payload alongside Image (Base64).