Status

Checks

• API and dashboardheadband.dev, the /v1 API and the dashboard, served by Vercel.Operational
• Control-plane databaseAccounts, projects, keys and usage. Checked by the API's health endpoint.Operational
• Engine server ewr1-aThe gateway in front of the hosted engines on ewr1-a, in New York (region ewr1).Operational
• Engine server ewr1-bThe gateway in front of the hosted engines on ewr1-b, in New York (region ewr1).Operational

What these checks mean

• The API check calls our public health endpoint, which also checks the database. An outage of the whole site may take this page down with it, since it is served by the same app.
• An engine server check calls that server's gateway from outside. A server that answers can still have one engine down, which this page does not show. Your dashboard shows the state of each of your engines.
• Each hosted engine has one home server. If its server goes down, an engine on bucket storage moves to the other server by itself in about 2 to 3 minutes. Always-on Meilisearch and Elasticsearch engines on bucket storage and a paid plan normally keep a live read-only copy on the other server, which answers their searches in the meantime. Any other engine is down until the server returns or we move it by hand. The SLA page explains why.
• Engines you bring are not checked here. The health endpoint is GET https://headband.dev/v1/health if you want to monitor it yourself.

Incident history

No incidents posted yet. This history starts with this page, on October 27, 2026. For anything affecting you now, email support@headband.dev.