For a few hours earlier today (from around 20:30 to 23:45 UTC), API actions were not finishing properly due to unexpectedly hitting plan limits for a related third-party service. I temporarily upgraded the plan to avoid further outages related to this, and have also reduced the usage as well.
Search daemons remained unaffected, and API actions would have completed on Flying Sphinx's side of things, but client calls would not have been alerted to this.
Apologies for the extended issue and belated report.