Errors & Rate Limits

All Presto APIs use standard HTTP status codes. Successful responses return 2xx; errors return the appropriate 4xx or 5xx code with a JSON body describing the problem.

The API enforces per-key rate limiting. Each API key has its own request budget, independent of other keys.

The default limit is 600 requests in any rolling 60-minute window per API key. This is a sliding window, not a fixed hourly reset — capacity frees up continuously as earlier requests age past 60 minutes. For sustained bulk submissions, pace your client at one request per 6 seconds or slower to stay under the limit indefinitely. Contact your organization administrator if you need a higher limit.

When you exceed the limit, the API returns 429 Too Many Requests with a Retry-After header indicating how many seconds to wait.

Every API response includes rate-limit headers following the IETF RateLimit header fields draft:

RateLimit-Policy: "default";q=600;w=3600
RateLimit: "default";r=277;t=42

On a 429, the standard Retry-After header (seconds) is also present:

HTTP/1.1 429 Too Many Requests
Retry-After: 42
RateLimit-Policy: "default";q=600;w=3600
RateLimit: "default";r=0;t=42

import re
import time

import requests

def submit(session: requests.Session, url: str, payload: dict) -> requests.Response:
    while True:
        resp = session.post(url, json=payload)
        if resp.status_code != 429:
            return resp
        wait = int(resp.headers.get("Retry-After", "60"))
        time.sleep(wait)

def remaining(resp: requests.Response) -> int | None:
    m = re.search(r"r=(\d+)", resp.headers.get("RateLimit", ""))
    return int(m.group(1)) if m else None

• 401/403: Do not retry — fix the authentication or authorization issue
• 429: Wait the number of seconds in the Retry-After header, then retry. If absent, use exponential backoff
• 500: Retry with backoff; if persistent, contact support