Predicts labels for a given text using the specified pretrained AI model (Preset).

This endpoint returns the AI model’s predicted labels for the provided text.

<b>language (optional)</b> — Two‑letter, lowercase ISO 639‑1 language code for the input text (e.g., "en", "tr", "es", "de", "fr"). Providing this lets Kimola skip automatic language detection and apply the most accurate prediction pipeline for that language. This is especially helpful for short texts or texts with typos. If omitted, Kimola handles language detection internally.

<b>aspectBased (optional)</b> — When true, analyzes the text by topics/aspects and returns multiple labels with their sentiment values (aspect‑based classification). When false or omitted, returns the dominant label prediction. <b>Note:</b> setting aspectBased=true consumes <b>2</b> queries (classification + sentiment), otherwise it consumes <b>1</b> query.

<b>text (required)</b> — The raw input text to classify.
If the AI model finds none of the labels relevant to the given text, it returns a special label None.