Vercel / Next.js threat watch

There are elevated rates of invocation failures for Hobby Team functions in new deployments. Existing deployments are unaffected. We are investigating. Mitigation is possible by rolling back to a previous deployment.

We are investigating elevated error rates for users logging in to Vercel using their GitHub account and connecting their GitHub accounts to Vercel. Login with email or other sign-in methods is unaffected.

We are investigating SAML Single Sign-On (SSO) errors, leading to inability to login using SAML/SSO or access SSO-protected teams through the Vercel dashboard and CLI.

We've identified an issue where some users may see missing Build CPU Minutes data on Usage pages in the Vercel Dashboard. The issue has been resolved, and we are backfilling the affected usage data.

We have identified an increase in erroneous 2FA (two-factor authentication) challenges and errors when accessing the Vercel dashboard and multiple API services for teams that require 2FA. We will share more information once we have it.

We've identified an issue where some customers may experience failed Builds beginning at 21:05 UTC. A rollback is in progress to mitigate this issue. We will provide additional updates as they become available. Failed deployments can be retried and should succeed.

We've identified an issue where some customers may receive erroneous spend management budget notifications due to a spend management budget miscalculation. Affected teams with spend management budgets configured to pause projects may also experience paused deployments. We are currently investigating this issue. We will provide additional updates as they become available.

The issue affecting Build Logs has been resolved. Between Jul 8 19:05 UTC and Jul 9 13:16 UTC, some Pro customers using standard machine types may have experienced missing Build Logs. New builds after the impact period are unaffected.

We've identified an issue where some customers may experience delays in builds starting and/or builds stuck in an initializing state. We are currently investigating this issue and will provide additional updates as they become available.

We are investigating an issue causing degraded performance across CDN, Dashboard, & Functions in Washington, Cleveland (iad1, cle1). We'll provide additional updates as they become available.

We've identified an issue where some workflow runs created between June 30 at 22:24 UTC and July 1 at 16:08 UTC may be stuck in a pending state or have reached a failed state. New workflow runs are not impacted. We are working on a fix for runs stuck in a pending state and will provide additional updates as they become available.

We are currently investigating the issue. Some users are experiencing errors logging in. The issue started at 11:30 AM UTC.

We are currently investigating an issue where some recently deployed Workflow projects are stuck as pending and not delivering queued messages as expected.

We're investigating an issue preventing customers from searching for and purchasing domains in the Vercel Dashboard.

Between Jun 19 09:09 and 16:16 UTC, a subset of Vercel Functions using node-fetch@2 may have experienced intermittent invocation errors that surfaced as ERR_STREAM_PREMATURE_CLOSE. As part of Node.js June 2026 security releases, we began rolling out new Node.js versions. Those versions contain an upstream regression that breaks response streaming for node-fetch@2, which caused the errors https://github.com/nodejs/node/issues/63989 We have resolved the issue by reverting to the previous Node.js version. No action is required — affected functions are now operating normally. We will re-land the Node.js upgrade once the upstream issue is fixed.

We are investigating a partial disruption to Observability, Web Analytics, Speed Analytics and Firewall features.

We are currently re-routing traffic from the FRA1 Vercel Region to the CDG1 Vercel Region. We are currently investigating this issue.

We are currently investigating this issue.

The issue has been identified and a fix is being implemented. A small number of requests may have seen elevated error rates in function invocations during this period.

We are investigating reports of some customers experiencing elevated errors creating new deployments. We will provide additional updates as they become available.