Amazon Bedrock (Runtime API) error
Amazon Bedrock 529 overloaded_error: what it means and how to fix it
529 overloaded_errorRetry? YesLast verified 4 documented fixes1 source
What it means
The model is temporarily unable to process the request because of high demand or insufficient serving capacity. This is a transient capacity error, different from a 429 ThrottlingException, which indicates the request exceeded an account quota.
Causes
- High demand for the model
- Insufficient serving capacity
How to fix it
- Fix 1.
Retry using exponential backoff and random jitter. If the response includes a Retry-After header, wait for the specified time before retrying.
- Fix 2.
Avoid immediate or synchronized retries from multiple clients.
- Fix 3.
If the model supports it, use cross-Region inference to route requests across multiple AWS Regions.
- Fix 4.
If the error persists, contact AWS Support with the request ID, model ID, AWS Region, and approximate timestamp.
More Amazon Bedrock errors
All 22 Amazon Bedrock errorsHTTP 529 at other providers
Compare 529 across providersFAQ
What does the Amazon Bedrock error "529 overloaded_error" mean?
The model is temporarily unable to process the request because of high demand or insufficient serving capacity. This is a transient capacity error, different from a 429 ThrottlingException, which indicates the request exceeded an account quota.
How do I fix Amazon Bedrock "529 overloaded_error"?
There are 4 documented fixes. Start with the first: Retry using exponential backoff and random jitter. If the response includes a Retry-After header, wait for the specified time before retrying.
Should I retry Amazon Bedrock "529 overloaded_error"?
Yes. The official docs treat this error as retryable.
Which other AI APIs return HTTP 529?
Per their official error docs: Anthropic API (529 overloaded_error).
Sources
Last verified against 1 source: Amazon Bedrock's official docs and official SDK repositories.
Newsletter
Know when the APIs change
One email per week. New model releases, API changes, what broke, and how developers are working around it.