Status | Rewind.ai

Checks run continuously. The timestamp on the page shows when the most recent check completed, and the page auto-refreshes every 60 seconds.

Operational means the service has no known issues, degradation or outages. Requests are completing within normal response-time targets.

Free-tier accounts do not carry a formal SLA, though they run on the same high-availability infrastructure. Paid plans include uptime commitments. Enterprise plans offer 99.9% or higher SLA with service credits if the target is missed.

Maintenance windows are scheduled during low-traffic periods and announced ahead of time. Rolling restarts keep most updates zero-downtime. Any exception is noted on this page in advance.

If a GPU server goes down, AI inference tools on that server become temporarily unavailable. The web app, account management and billing continue to operate. Recovery is prioritised and updates are posted here as they happen.

Yes. Database backups run on a regular schedule and are encrypted at rest. Both the web server and GPU servers have built-in redundancy. Enterprise customers can request multi-region deployments for additional coverage.

Enterprise customers receive direct outage and maintenance notifications. A broader notification system for all users is in progress. In the meantime, check this page or contact us if you run into an issue.

Systems are monitored around the clock. Full outages are usually detected within minutes and resolved within hours. Degraded-performance events are investigated immediately and progress is communicated on this page.

PostgreSQL stores user accounts, token balances, chat history, API keys, payment records and usage statistics. AI model weights and generated content are not stored in the database long-term.