Google said Wednesday that its Gemini service was mostly restored after a disruption that pushed downdetector reports above 1,600 and produced repeated error codes for users.
The company confirmed at 6 p.m. PT (9 p.m. ET) that the majority of affected users should be able to use Gemini again after engineers applied mitigations. The problem began at 6:26 a.m. ET and produced error codes 1099 and 1076 with the message, "Something went wrong," on Gemini inside Google Workspace and in the Gemini app.
Reports peaked at more than 1,600 during morning work hours, affecting the Gemini integration in Google Workspace — including Drive, Docs, Sheets and Slides — as well as the standalone Gemini app on web, MacOS, iOS, Chrome and Android. Google’s status page classified the incident as a service disruption and noted no workaround was available at the time of its update.
Google said the disruption was triggered by a performance issue in a backend database that impeded retrieval of the Gemini App tools catalog. Engineers addressed the immediate problem by optimizing load distribution across that backend database; the company said those fixes removed the impact for most users and that its team would continue to monitor the service.
The company first posted an official status update at 8:40 p.m. ET and scheduled follow-ups, telling users to expect another update at 8:30 p.m. PT and in a separate update by 3:30 p.m. ET. Even as Google reported the service was mostly resolved, the Workspace Status Dashboard still showed the disruption as ongoing — a discrepancy that left some customers seeing the service as healthy while the official dashboard had not yet cleared the incident.
Users encountering the disruption saw the same generic failure codes across platforms; Google’s explanation tied those failures to the tools-catalog retrieval problem inside its backend. Earlier in the day, initial spikes of reports were tracked in real time: an earlier FilmoGaz update captured a rise to 353 reports within the first hour of the incident ( and other outage coverage has compared surge patterns to previous service interruptions (
Google said its engineering team applied mitigations and optimized backend load distribution to restore service, but did not list any longer-term preventive measures or configuration changes meant to stop a similar database performance issue from recurring. The company said it would post another update at the announced times while monitoring performance.
The immediate customer impact now appears limited: most users can access Gemini in Workspace and the Gemini app across devices. The unresolved question is operational and specific — will Google publish the concrete steps it will take to prevent a repeat of a backend database performance failure that pulled down Gemini across multiple platforms?
