Skip to main content

Rate Limiting

To guarantee low audio latency and 99.99% uptime across our real-time voice infrastructure, Sayvy AI enforces rate limits on REST API requests and concurrent voice audio sessions.

Tier Rate Limits


Rate Limit Headers

Every HTTP response from Sayvy AI includes headers describing your current quota status:

Handling 429 Responses

When rate limits are exceeded, Sayvy AI responds with HTTP status code 429 Too Many Requests:

Implementing Exponential Backoff with Jitter

To reliably recover from rate limits, implement exponential backoff with full jitter in your client code:
If your workload requires higher outbound call concurrency or custom burst allowances, contact our team at [email protected].