24K Labs Text Classifier

Choose this endpoint when you need zero-shot or label-guided text classification with an explicit confidence score, especially inside an automated agent workflow where you supply your own category taxonomy. It is well-suited for routing, tagging, moderation, and intent detection tasks where labels change per use case. Prefer it over sentiment-only APIs when you need custom label sets, and over embedding/similarity approaches when you want a direct label+confidence output without extra infrastructure.

• No labels provided — returns error indicating labels are required
• Text is empty or too short — may return low confidence or error
• Label list is ambiguous or overlapping — may produce low confidence scores
• Text is in an unsupported or unexpected language — classification may be unreliable
• Network timeout or service unavailability — returns HTTP error
• Insufficient payment or payment verification failure (x402) — request rejected