ChatGPTOpenAI
OperationalNotice Elevated error rates for most hosted Code Reviews and Security Reviews on private GitHub repositories
Uptime · 90/90 days99.1%
A live status board for the major AI services. Every 5 minutes we check the official status feeds of ChatGPT, Claude, Gemini & co. – so in seconds you know whether it's you or the provider.
Notice Elevated error rates for most hosted Code Reviews and Security Reviews on private GitHub repositories
Each bar is one of the last 90 days: green = up, amber = degraded, red = outage. The window moves forward every day.
Measured uptime over the last 90 days – the higher, the more stable the service. Sampled every 5 minutes from the providers' official status feeds.
Current and past incidents across the monitored AI services – newest first.
Elevated error rates for most hosted Code Reviews and Security Reviews on private GitHub repositories
Ads POST /ads/* endpoints elevated error rates
Elevated errors for Claude Fable 5.1
Elevated errors for Claude Sonnet 5.5 and other models
Elevated errors for Claude Fable 5 and Claude Fable 5.1
Elevated errors for Claude Opus 5.5
Grok 4.6 and Grok 4.7 Service Issues
Elevated errors for Claude Fable 5.1, Sonnet 5.5, and Haiku 5.5
Degraded experience with ChatGPT app connections using CIMD
Help Center Chat Support Degraded
Delays in customer support responses
New Ad Accounts failing to provision admins
Degraded Performance of Dots on Enterprise
Elevated errors affecting Connectors
Elevated ChatGPT conversation and Responses API errors
Degraded performance for Claude Fable 5
Elevated error rates for gpt-5.6-luna API requests
Elevated FedRAMP API error rates for Astra and Sol
Degraded performance in ChatGPT, Codex, and APIs
File Download errors
Widespread feature degradation
Delayed Costs data in the Compliance API
Limited availability on gpt-image-2.5-flare and gpt-image-2.5-sunburst
Degraded performance in Codex and ChatGPT Work
Multiple US endpoints experiencing timeouts and latency
Degraded performance in GPT-6.1 Sol Codex and Work mode
Model Performance or Availability Issue
GroqCloud Console Errors Affecting Multiple Regions
Model Performance or Availability Issue - minimaxai/minimax-m2.7
Workspace admins may be unable to manage plugins
Some ChatGPT content may be unavailable on Android
Delayed Costs data in the Compliance API
Increased Dubbing Latency and Failure Rate
Agent SIP Conversation Disruption
Music 2.5 endpoints performance degraded
Some APAC users are seeing errors in ChatGPT Work, conversations, and GPTs.
Elevated errors with GPT-5.6 Instant in FedRAMP workspaces
Degraded performance in Codex and Work mode: Dot turn failures, delayed responses, and reduced proactivity
Dot in Codex may fail to create new threads
Elevated errors on platform.claude.com
Honest and verifiable: we only query the providers' official, public status feeds – no private logins, no guesswork.
A service samples each provider's official status feed on a 5-minute cycle – whether it's Atlassian Statuspage, Instatus or Google Cloud.
From those samples we compute uptime per service per day and carry it forward. A 90-day history can only ever be built going forward.
Active reports are picked up and automatically closed once they leave every feed – giving a continuous record.
If a status feed can't be read, we mark that service as "unknown" – it never counts as an outage.
The list grows. More providers such as Mistral and Grok will follow as soon as they offer a machine-readable status feed.
The board above shows the live status of ChatGPT (OpenAI) and every other service. "Operational" means no incident reported. "Degraded", "Partial outage" or "Outage" means the provider is currently reporting problems. "Maintenance" means planned work. The board refreshes automatically every minute.
Exclusively from the providers' official, public status pages (e.g. status.openai.com, status.claude.com). We read them every 5 minutes and normalise them into one view – we are not affiliated with any of the providers.
Our service checks the feeds every 5 minutes; the page additionally pulls the latest state every minute. The 90-day uptime is carried forward daily.
We start with eight major services that offer a machine-readable status feed. Providers without an open status feed (currently e.g. Mistral) are added as soon as that changes.
Each bar is one day: green = up, amber = degraded, red = outage. The bar shows the last 90 days and moves forward every day. It is built from our own samples, so it can't be back-filled.
This board summarises the providers' official status reports and is for information only. i6eal is not affiliated with the named providers; the authoritative source is always each provider's own status page. Brand and product names belong to their respective owners.
We build AI solutions for mid-sized companies with resilience, fallback models and monitoring – so your operations stay stable.
These tools cover related ground.