Published
TypeSafe posted an API and console outage, then said it was back
At 08:30 UTC on September 21, TypeSafe posted that the API and console had issues and pointed at status.typesafe.ai. At 09:53 UTC it posted "We are back!" The status incident is titled "Console is unavailable" and marked resolved at 08:16 UTC, before those tweets.
TypeSafe posted at 08:30 UTC on September 21 that it was seeing issues on the API and console, and that it was investigating. The link is status.typesafe.ai. At 09:53 UTC the same account posted “We are back!” quoting that notice.
The status page, when we read it later that morning, showed all services online. Last updated September 21, 2026 at 10:38am UTC. api.typesafe.ai listed 99.841% uptime. console.typesafe.ai listed 99.988%.
The named incident is Console is unavailable. It lists the console as the affected service and is marked resolved September 21, 2026 at 8:16am UTC, with the sentence “Issues with TypeSafe console and API are fully resolved.” That resolve time is earlier than the 08:30 investigating tweet. The incident page does not publish a start time or a duration.
The 90-day API history also records 18 minutes of downtime on September 19 and 5 minutes on September 16.
We did not send a request during the window.
This site's reading
Editorial notes evaluating claims against primary sources, contextualizing findings alongside related implementations, and defining technical terms.
Verify
The September 21 08:30 UTC post from @typesafeai is the primary source: "We are seeing some issues on our API/console, we are investigating," with a link to status.typesafe.ai. The 09:53 UTC follow-up is "We are back!" quoting that post. The status page, when we read it, showed all services online, last updated Sep 21, 2026 at 10:38am UTC. api.typesafe.ai listed 99.841% uptime. console.typesafe.ai listed 99.988%. The named incident is "Console is unavailable," affecting console.typesafe.ai, marked resolved Sep 21, 2026 at 8:16am UTC, with the line "Issues with TypeSafe console and API are fully resolved." That resolve timestamp sits before the 08:30 investigating tweet. The incident page does not publish a start time or a duration. The 90-day API history also shows 18 minutes of downtime on September 19 and 5 minutes on September 16. We did not send a request during the window.
Compare
The waitlist story on this desk covered how to get a key. LiteLLM's typesafe compaction guardrail fails open if Jev is down. term-llm's control plane also fails open to the main model. This incident is the first TypeSafe-posted API/console outage in the digest. The status page's resolve time and the two tweets do not line up, so the duration is whatever those three pages each say.
Terms
- status.typesafe.ai
- TypeSafe's public status page for api.typesafe.ai and console.typesafe.ai. The September 21 incident is titled Console is unavailable.