DEVELOPER GUIDE

Laya AI API

Send a short state and typed questions to POST /api/v1/systemone. The response contains the selected model, answers, and token usage. Requests require a Laya AI API key and enough input tokens or a welcome credit. The API uses purchased tokens first, then one credit for the whole request. A disconnected GPU service returns 503 before charging.

curl https://laya-ai.pro/api/v1/systemone \
  -H "Authorization: Bearer $LAYA_AI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"laya-english","state":"The checkout page fails after card entry.","questions":{"team":{"type":"choice","instructions":"Who should handle this?","criteria":{"billing":null,"engineering":null,"support":null}}}}'

Model choices

laya-english uses a 512-token budget per question. laya-multilingual uses 1024. Both budgets include the text, question, and labels. For longer input, set model to jev-latest (64k tokens) or mercury-decide (32,768 tokens) with the same request. Oversized input returns an error instead of a shortened decision. Compare the models.