Between approximately 10:20pm and 11:15pm EST, the API experienced slow responses and increased errors. Service has been restored and is stable. We are investigating the cause and working to prevent this from happening again.