25,000 item execution history in a workflow. This means that your workflow can’t have more than 25,000 state transitions in a single execution. For most use cases, this won’t be an issue. If you do have very long*running executions that may reach such a high number of state transitions, you’ll need to consider splitting your workflow into multiple workflows to stay under the 25,000 transitions limit.

1MB maximum request size. A request to AWS Step Functions can’t have a payload larger than 1MB. If you’d like to use larger files as inputs to a Step Functions workflow, consider using Amazon S3 to store the files and use the S3 URIs as inputs to further operations.

Spikes in AWS API requests caused by a workflow. AWS may throttle AWS API requests that come in spikes. A sustained rate of requests will cause no trouble, but if some parts of your workflow use the AWS API inefficiently, a sudden flood of requests from your workflow might trigger the API limits. You can work around this by, if possible, grouping requests to the same AWS service into a single API call, or by introducing timeouts between operations.

50 tags per resource. You can’t apply more than 50 tags to any of the AWS Step Functions resources. If this presents an issue for you, consider changing your tagging scheme to stay under the limit.

https://www.serverless.com/guides/aws-step-functions