Rate Limits

The per key sliding window rate limit, the 429 response it produces and how to back off.

Showing instructions for Xebit VoiceCloud.Showing instructions for Xebit VoiceKit, on your own infrastructure. Switch with the selector at the top. Updated 22 October 2026.

Each API key is rate limited with a sliding window to keep the service stable for everyone. When you go over, the API answers 429 and tells you how long to wait.

API throttling

  • Threshold: 60 requests per minute (RPM) per API key.
  • Threshold: configured by your deployment. Xebit does not impose a figure on Xebit VoiceKit.
  • Scope: all standard v1 endpoints.
  • Enforcement: a sliding window calculated at the API gateway.

On Xebit VoiceKit, Xebit does not impose these figures. Limits are set by your own deployment: the infrastructure you run it on and your licence band. Your platform team sets them as part of the deployment.

Handling throttling

When you exceed the limit the API responds with:

  • Status: 429 Too Many Requests
  • Header: Retry-After, the whole number of seconds to wait before retrying
  • Body:
JSON
{"error": {"code": "RATE_LIMIT_EXCEEDED", "message": "..."}}

Retry with exponential backoff, honouring Retry-After when it is present, so a burst of 429s does not turn into more load on the gateway.

Backoff example

# -i prints the headers so you can read Retry-After on a 429
curl -s -i "https://api.voice.xebit.ai/v1/calls/call_abc123" \
  -H "x-api-key: $XEBIT_API_KEY"
# -i prints the headers so you can read Retry-After on a 429
curl -s -i "https://<your-xebit-voicekit-host>/v1/calls/call_abc123" \
  -H "x-api-key: $XEBIT_API_KEY"