API

Base address https://promptlift.aiyakala.com/api/v1. Send an API key from your dashboard as Authorization: Bearer <key>. Each key may make 30 requests a minute.

Score a prompt

curl -sS https://promptlift.aiyakala.com/api/v1/evaluate \
  -H "Authorization: Bearer $PROMPTLIFT_KEY" -H "Content-Type: application/json" \
  -d '{"prompt": "Write a summary of the attached report for our sales team.", "improve": false}'

prompt: up to 20,000 characters, not stored (only its SHA-256). improve: also return a rewritten prompt; uses more tokens. The answer has score (0–100), criteria (goal, context, constraints, efficiency, risk; each with score, max, reason, suggestion), usage.tokens and quota.

Check a text's AI style

curl -sS https://promptlift.aiyakala.com/api/v1/style -H "Authorization: Bearer $PROMPTLIFT_KEY" \
  -H "Content-Type: application/json" -d '{"text": "..."}'

Up to 100,000 characters. Returns a 0–100 estimate of how machine-written the text reads and what it found. It is an estimate of style, not proof of who wrote it, and costs a tenth of the text's tokens.

Usage

GET https://promptlift.aiyakala.com/api/v1/usage returns your plan, this 4-hour window and this month. Windows start at 00:00, 04:00, 08:00, 12:00, 16:00 and 20:00 UTC.

Errors

Status	Code	Meaning
400	`invalid_request, empty, too_long`	The body is wrong or the input is too long.
401	`unauthorized`	No valid API key.
402	`window_quota, monthly_quota, payg_limit`	Quota used up; retry_after says when it refreshes.
403	`suspended`	The account is suspended.
429	`rate_limited, failure_limit`	Over 30 requests a minute, or too many failed requests in an hour.
502	`bad_model_output`	The model's answer was unusable. Not charged.
503	`global_ceiling, model_unavailable`	At capacity or the model is down. Not charged.