Datvero automation observatory · status.openai.com
No. OpenAI's status page reports all systems operational for ChatGPT and the OpenAI API: Conversations, Chat Completions, Responses and 21 other components (checked November 6, 2026, 18:10 UTC; this page refreshes every 5 minutes).
Read from OpenAI's official status API at November 6, 2026, 18:10 UTC. Datvero republishes what OpenAI states; it is not an independent uptime measurement.
From October 7 to November 6, 2026, status.openai.com lists 30 incidents: 0 critical, 4 major, 23 minor and 3 with no impact (the vendor's own classification). Median time from first post to resolution, impact incidents only: 2 h. Longest: “Delayed Costs data in the Compliance API” (October 9, 2026), 2.9 days. In the last 30 days: 29 incidents, 4 of them major or critical.
Retry with a growing delay, send failures to an error workflow, and get alerted when runs stop succeeding. In n8n, every node has a Retry On Fail setting; Make has Break error handlers that retry incomplete executions; Zapier can replay failed runs with Autoreplay. Datvero watches the runs of your n8n, Make and Zapier workflows and alerts you when they fail, whatever the cause.
No. OpenAI's status page reports all systems operational for ChatGPT and the OpenAI API: Conversations, Chat Completions, Responses and 21 other components (checked November 6, 2026, 18:10 UTC; this page refreshes every 5 minutes).
From October 7 to November 6, 2026, status.openai.com lists 30 incidents: 0 critical, 4 major, 23 minor and 3 with no impact (the vendor's own classification). Median time from first post to resolution, impact incidents only: 2 h. Longest: “Delayed Costs data in the Compliance API” (October 9, 2026), 2.9 days. In the last 30 days: 29 incidents, 4 of them major or critical.
OpenAI publishes it at https://status.openai.com/. This page republishes its public feed every 5 minutes with timestamps. Datvero is not affiliated with OpenAI.
The status page lists what OpenAI has declared. If it is green and you still see errors, read the error code, retry with a growing delay, and check whether the problem is specific to your account, your region or one integration. A green page does not prove that every request to ChatGPT succeeds.
OpenAI’s public status feed does not attach incidents to components, and it lists no component named Codex (only “Codex in ChatGPT Desktop”). A separate page would have no data of its own, so this page covers ChatGPT and the OpenAI API together.
Retry with a growing delay, send failures to an error workflow, and get alerted when runs stop succeeding. In n8n, every node has a Retry On Fail setting; Make has Break error handlers that retry incomplete executions; Zapier can replay failed runs with Autoreplay. Datvero watches the runs of your n8n, Make and Zapier workflows and alerts you when they fail, whatever the cause.
Method: OpenAI's public status API (status.openai.com/api/v2) is read every 5 minutes, its incident list every 15 minutes, and each incident is kept with its official timestamps. Durations run from the first post to the resolved post. Machine readable copy: services.json. Official source: https://status.openai.com/.
Statuses, incident names and texts belong to OpenAI. Datvero is not affiliated with OpenAI. A green status page does not mean every workflow that calls ChatGPT succeeds.