# Rate limits

Rate limiting controls the frequency at which requests are made to API endpoints within specific time periods.

It helps protect against service overload while ensuring consistent performance for all clients. Exceeding limits results in a `429 Too many requests` error response.

**API usage limits**

Solidgate returns the `429` error response when necessary to protect legitimate merchant traffic.

Rate limits differ by endpoint based on operational and reliability needs. The Solidgate team continuously monitors system performance and may adjust these limits as needed to maintain optimal service quality.

For endpoint-specific rate limit information, visit the **Developers** section in the Solidgate Hub, which is updated as changes occur.

**Handle rate limits**

You can handle rate limiting by monitoring for the `429 Too many requests` error response. Effective handling combines retries and overall request flow control.

A widely used approach for handling rate limit error responses is implementing exponential backoff with jitter. This method retries requests using short initial delays that increase after each failure. Introducing randomization, or jitter, helps avoid conflicts caused by multiple clients retrying simultaneously.

While retries are useful, a significant improvement comes from regulating request flow across the entire application. The token bucket is standard practice for this purpose. It allows short bursts of requests while enforcing an average request rate over time, reducing traffic spikes and improving overall stability.

For help, contact us.