String docs

Rate limits and concurrency

Request rate limits, concurrent fetch and browser limits, and how to handle 429s.

Requests are rate limited per organization. When you exceed the limit, requests return 429 Too Many Requests. Free and Starter get 15 requests per second, Growth 30, Scale 60 and Volume 90, and a higher rate is available on request.

Concurrency

Every plan can have up to 10,000 fetches and 5,000 browser sessions in flight at once. You do not need to reserve or buy parallel slots.

In-flight requests settle at roughly your request rate multiplied by your average response time. At 60 requests per second with a 0.5 s average response, that is about 30 requests in flight; if responses average 2 s, about 120. Keep the average rate under the limit below and the parallelism takes care of itself.

Rate limit

PlanRequest rate
Free15 / sec
Starter15 / sec
Growth30 / sec
Scale60+ / sec
Volume90+ / sec
EnterpriseCustom

Your request allowance replenishes continuously at your plan's request rate, up to a maximum burst capacity. Each admitted request uses one unit of the allowance.

At the Free rate, an organization using its signup credit can:

  • Start up to 60 requests immediately when its allowance is full.
  • Replenish its allowance at 15 requests per second, up to the 60-request maximum.
  • Refill an empty allowance completely after 4 seconds without requests.

You can use the initial burst and continue at 15 requests per second as the allowance replenishes. The four seconds is the full refill time, not a cap of 60 requests in every rolling four-second window. Sustained traffic above your plan's rate eventually exhausts the allowance and gets 429s.

Once your organization subscribes or adds funds, the default burst capacity is 60 seconds of its request rate.

Until your organization subscribes to a plan, it is on the Free rate. Subscribing to Growth or above raises it within about a minute, with nothing to change in your integration.

Need more? Email [email protected] with your expected request volume. A raised limit is applied to your organization and takes effect on your next request — there is nothing to change in your integration.

Rate limit headers

The API reports the applicable request-rate allowance for /fetch, /search, sitemap jobs, contexts and sessions in the IETF RateLimit header fields:

RateLimit-Policy: "burst";q=3600;w=60
RateLimit: "burst";r=3412

RateLimit-Policy reports the maximum burst capacity (q, requests) and the time to refill an empty allowance (w, seconds). The sustained request rate is q / w: the example above permits a burst of 3,600 requests and replenishes at 60 requests per second. An organization using signup credit gets q=60;w=4 at the Free rate.

RateLimit reports how much of the allowance is left (r). It replenishes continuously, so the header carries no reset time. Slow down as r approaches 0.

These headers may be absent. If they are, pace requests at your organization's documented rate and follow Handling 429s if a request returns 429.

Requests without a valid API key

When /fetch or /search accepts a request without an API key, the response includes X-String-Keyless: true. Its RateLimit headers describe the applicable request-rate allowance. The daily keyless allowance is separate; a daily-limit 429 uses Retry-After to tell you when it resets.

A request with an invalid API key returns 401 without an organization's rate-limit headers. Endpoints that require an API key also return 401 when the key is missing. Check your API key before retrying a 401.

429 response

{
  "error": "Rate limit exceeded",
  "message": "Reduce request volume to 60 requests per second on average.",
  "limitRequestsPerSecond": 60
}

The response includes an X-RateLimit-Limit-Rps header with your allowed average requests per second, and Retry-After: 1. That header and limitRequestsPerSecond both report your organization's limit.

A 429 with errorType: "TargetFailingError" or errorType: "TargetRequestLimitError" is about one destination, not your limit, and a 502 with errorType: "OriginRateLimitedError" is the destination's own rate limit. See A 429 about the destination and Destination rate limiting.

Handling 429s

  • Back off and retry with exponential backoff and jitter.
  • Smooth bursts on the client side so your average stays under the limit.
  • Cache results where you can to avoid re-fetching the same URL.
  • Throttle on rate, not on parallelism. Lowering your worker count is an indirect way to slow the request rate and costs you latency; a rate limiter on the client is the direct fix.