Retries

Configure Helicone to automatically retry failed LLM requests, overcoming rate limits and server issues using intelligent exponential backoff.

Who can use this feature: Anyone on any plan.

Introduction

Retrying requests is a common best practice when dealing with overloaded servers or hitting rate limits. These issues typically manifest as HTTP status codes 429 (Too Many Requests) and 500 (Internal Server Error).

For more information on error codes, see the OpenAI API error codes documentation.

Learn About Exponential Backoff

To effectively deal with retries, we use a strategy called exponential backoff. Exponential backoff involves increasing the wait time between retries exponentially, which helps to spread out the request load and gives the server a chance to recover. This is done by multiplying the wait time by a factor (default is 2) for each subsequent retry.

Why Retries

• Overcoming rate limits and server overload.
• Reducing the load on the server, increasing the likelihood of request success on subsequent attempts.

Quick Start

To get started, set Helicone-Retry-Enabled to true.

When a retry happens, the request will be logged in Helicone.

Retries Parameters

You can customize the behavior of the retries feature by setting additional headers in your request.

Parameter	Description
helicone-retry-num	Number of retries
helicone-retry-factor	The exponential backoff factor used to increaase the wait time between subsequent retries. The default is usually 2.
helicone-retry-min-timeout	Minimum timeout (in milliseconds) between retries
helicone-retry-max-timeout	Maximum timeout (in milliseconds) between retries

Header values have to be strings. For example, "helicone-retry-num": "3".