GitHub promises 99.9% uptime in its Enterprise Cloud SLA — roughly 8.7 hours of allowable downtime per year. Based on incident reports and status page data, the platform has consistently fallen short of that target across 2025 and into 2026.
GitHub Uptime Summary
| Period | Uptime | SLA Target | Status |
|---|---|---|---|
| 2025 (worst point) | <90% | 99.9% | SLA missed |
| February 2026 | Below 99.9% | 99.9% | SLA missed |
| Feb 9–10, 2026 | ~95% estimated | 99.9% | Major outage |
What 99.9% Uptime Means in Practice
- 99.9% = max 8.7 hours downtime/year, ~43 minutes/month
- 99.5% = max 43.8 hours downtime/year, ~3.6 hours/month
- 99.0% = max 87.6 hours downtime/year, ~7.3 hours/month
- 90% = max 876 hours downtime/year — nearly 37 days offline annually
GitHub’s unofficial 2025 uptime dipping below 90% at one point represents availability worse than a service that is effectively down for over a month per year.
February 2026 Major Incident Timeline
- Feb 9, 1554 UTC: GitHub acknowledged problems with core services
- 1629 UTC: Copilot began experiencing extended issues
- 1757 UTC: Notification delays improved to ~30 minutes but unresolved
- 1929 UTC: Core services partially restored
- Feb 10, 0957 UTC: Copilot issues resolved — 16+ hours after they began
Services affected: GitHub Actions, pull requests, notifications, Copilot, and general repository operations. For full coverage of this incident, see our GitHub outages analysis.
What GitHub’s SLA Actually Covers
GitHub’s 99.9% uptime SLA applies only to Enterprise Cloud customers. Free and Team plan users have no contractual uptime guarantee. Even Enterprise customers must proactively file for SLA credits — compensation is not automatic.
GitHub vs Alternatives: Published SLA Comparison
| Platform | Published SLA | Notes |
|---|---|---|
| GitHub Enterprise Cloud | 99.9% | Enterprise only; repeatedly missed |
| GitLab.com | 99.95% | Higher published target |
| Bitbucket Cloud | 99.9% | Atlassian infrastructure |
| Azure DevOps | 99.9% | Same Microsoft infrastructure as GitHub |
How to Check GitHub Uptime Yourself
GitHub’s official status page is at githubstatus.com with a rolling 90-day view. The format was changed in 2025 making precise uptime calculations harder. For independent monitoring:
- UptimeRobot — free tier monitors github.com every 5 minutes
- GitHub’s API: poll
https://api.github.comas a health check - RSS feed: GitHub publishes an incident RSS feed via statuspage.io
Protecting Your Team from GitHub Outages
- Mirror critical repositories to a secondary host (GitLab, self-hosted Gitea)
- Cache CI dependencies so pipelines continue when Actions is degraded
- Monitor GitHub status proactively rather than waiting for developer reports
- Consider self-hosted runners on separate infrastructure for critical deployments
Bottom Line
GitHub’s reliability record in 2025-2026 shows a consistent pattern of missing its own 99.9% SLA. Teams that treat GitHub as fully reliable infrastructure with no fallback plan are exposed every time an outage hits. The frequency and duration of recent incidents suggests the gap between SLA promise and delivered reliability is not improving.
