GitHub Copilot ran into elevated errors on its GPT-5.4 Nano model for 38 minutes on 2 November 2026. GitHub traced the fault to an upstream model provider, not to its own infrastructure.
Started
17:46 UTC
Duration
Lasted 38m
Source
IsDown
Next time
GitHub is degraded. Is your product working?
An incident on their side does not always mean an outage on yours, and the only way to know is to already be watching your own endpoints. Set that up in under a minute and you will be notified the next time your own site is affected. And there will be a next time.
50 monitors for your own websites. No card.
What happened?
GitHub's status page opened an incident at 17:46 UTC titled "Elevated errors on GPT-5.4 Nano". The first two updates, both posted at 17:46, describe the same problem from two angles: degraded performance for Copilot AI Model Providers generally, and elevated errors specifically on the GPT-5.4 Nano model. At 17:58 GitHub added detail, saying availability was degraded for GPT-5.4 Nano "in Copilot products and IDE surfaces" and attributing it to "an issue with an upstream model provider". That single line is the only explanation GitHub has given for why this happened, and it points outside GitHub's own systems rather than at them. By 18:10 GitHub said the degradation had been mitigated and moved to monitoring. At 18:24 the incident was marked resolved, with GitHub promising a root cause analysis "as soon as it is available", which has not appeared yet. The whole incident ran from 17:46 to 18:24, 38 minutes in total. No components are listed as affected on GitHub's own breakdown for this incident, and no user reports were filed during the window. GitHub's 90 day history shows 61 incidents with a median duration of 1 hour 26 minutes, so this one resolved faster than GitHub's own typical pace. Across GitHub's full history, 1,026 incidents have been logged at a rate of 13.2 per month.
Learning
UptimeRobot's own checks against GitHub's api-auth and raw-cdn components, run every 60 seconds from both the EU and North America, stayed at 100% uptime across the 24 hour, 7 day and 30 day windows, and this incident was never registered against those probes. That is not a contradiction. The fault sat in a Copilot AI feature calling out to an upstream model provider, a layer that a check against authentication or CDN endpoints was never built to see. This is the general shape of AI-feature outages: the core platform keeps answering normally while a specific model integration degrades underneath it, invisible to anything watching the front door. If you only watch your own endpoints, you would have seen nothing wrong either, because the break happened one layer further out, between GitHub and its model provider. Over the last 30 days UptimeRobot logged 21 incidents on GitHub, all resolved, with a median duration of 1h 16m, which gives a sense of how often this kind of gap between "platform is up" and "feature is degraded" opens. Your monitoring covers your half, UptimeRobot now watches the provider's half too, so next time you know which side broke without guessing.