Errors & Rate Limits

HTTP status codes

Code	Meaning
`200` / `201` / `202`	Success. `202` means accepted for background processing.
`401`	Missing or invalid API key.
`403`	Browser key used from a domain that isn't on its allow-list.
`404`	The referenced `request_id` doesn't exist on your account.
`422`	Request is missing a required field or failed validation. Check `error` / `details` in the response body.
`429`	Rate limit exceeded. See below.
`500`	Something went wrong on our end. The response still includes `request_id` — send it to us if you need to investigate.

Error response shape

{
  "error": "Invalid request",
  "details": [
    "Email can't be blank"
  ],
  "request_id": "550e8400-e29b-41d4-a716-446655440000"
}

Rate limits

Your account has one per-minute limit, shared across every server API key on it and across every server you call us from — neither creating more keys nor adding more machines raises it. Each browser key has its own separate limit, counted per visitor. Limits are configured per account — ask your Revylta contact if you need yours raised for a launch or a traffic spike.

`429 Too Many Requests`	Returned once you exceed your limit.
`Retry-After` header	Seconds to wait before retrying. Only sent on a `429`.

Every /api/v1 response — not just 429s — carries three headers so your integration can pace itself before ever being blocked:

`RateLimit-Limit`	Your per-minute limit for the endpoint you called.
`RateLimit-Remaining`	Requests left in the current one-minute window.
`RateLimit-Reset`	Seconds until the window resets.

Request IDs

Every score response carries a request_id. Generate your own UUID and send it as the X-Request-ID header if you want to control it (useful for idempotency/retries); otherwise one is generated for you. Keep it — it's how you fetch background results, send telemetry, and report outcomes for the same request later.