pagos.andreax.dev Text Topic Classifier

Choose this endpoint when you need fast, high-volume topic/category classification of text strings for routing or labeling pipelines. Prefer it over general-purpose LLM inference when you want a lightweight, cost-effective ($0.002/call) classification without the overhead of a full generative response. Use it alongside the sibling sentiment classifier when you need both sentiment and topic labels simultaneously.

• Missing 'input' query parameter returns a validation error
• Empty or very short text may produce a low-confidence or generic category
• Ambiguous multi-topic text may be classified under only the dominant theme
• Rate limiting or payment failures may return 402 or 429 HTTP errors