• 228 destinations
  • No KYC, ever
  • Bitcoin, Monero, USDT and 4 more

API and integration

Rate limit

A rate limit is the maximum number of API requests accepted per second (or minute) per key; above it the API answers 429 with a Retry-After header, and batch endpoints are the way to send volume.

A rate limit caps how many requests one credential may make in a period, usually a second or a minute. It protects the platform from runaway loops and each account from the side effects of someone else's bug. Implementations vary: fixed windows count requests per clock interval, sliding windows look back over a moving one, and token buckets allow short bursts while holding the average. The status code for "too many" is 429; a well-behaved API says how long to wait in a Retry-After header and exposes the remaining budget in response headers, so that clients can slow down before they hit the limit.

Rate limits are about requests, not messages. One request carrying a thousand messages counts once, which is why volume belongs in batch calls rather than in many parallel single sends. Nor is the API limit the same thing as throughput, the pace at which carriers deliver, or pacing, the speed you choose for a campaign.

Example

A hypothetical sign-up service sends each code in its own request. During a promotion, sign-ups spike and the service starts receiving 429 responses. The fix is not more keys or more threads: codes go through a small queue that honours Retry-After, and the newsletter that was going out at the same moment moves to batch requests.

Common mistakes

  • Retrying a 429 immediately, in a tight loop, which keeps the client over the limit.
  • Ignoring the remaining-budget header until requests start failing.
  • Treating a 429 as a failed message; nothing was sent, and a later retry with the same idempotency key is safe.
  • Spreading traffic across extra keys to dodge the limit instead of batching.

At SMSMeteor: the limit is 50 requests per second per key, and a batch of up to 1,000 messages counts as one request. Responses carry X-RateLimit-Limit and X-RateLimit-Remaining, and a 429 carries Retry-After. Because the limit is per key, one key per system also stops a runaway job in one integration from using up the budget of another.

Go further

See the term in practice.An email address opens the account, the rate card is public, payment is crypto, the test key is free.
Start sending