Is AssemblyAI Down?

No — AssemblyAI is up

AssemblyAI's status page: Operational · no open incidents
FlareWarden checks: Website reachable · 8 of 8 regions · 212 ms

Right now

The verdict is the worse of these two. How we decide

AssemblyAI's status page

status.assemblyai.com · read 3 minutes ago · checked every 3 minutes

Operational

AssemblyAI's components, as its status page reports them
Component	Status
Asynchronous API	Operational
Streaming API	Operational
LLM Gateway	Operational
Transcription Queue	Operational
Container Registry	Operational
Compute Instances	Operational
Container Control Plane	Operational
Load Balancers	Operational
Database	Operational
Usage Statistics Database	Operational
Object Storage	Operational
Usage Statistics Bus	Operational

4 more, all operational · show all 16

All of AssemblyAI's components
Component	Status
APIs
Asynchronous API	Operational
Streaming API	Operational
LLM Gateway	Operational
AWS Cloud
Transcription Queue	Operational
Container Registry	Operational
Compute Instances	Operational
Container Control Plane	Operational
Load Balancers	Operational
Database	Operational
Usage Statistics Database	Operational
Object Storage	Operational
Usage Statistics Bus	Operational
Web
Website	Operational
Playground	Operational
Dashboard	Operational
Billing	Operational

Open their status page

FlareWarden checks

assemblyai.com from 8 regions · when you open this page (at most every 30 minutes), otherwise every 6 hours

Reachable

Regions up: 8 of 8
Response: 212 ms
Checked: 4 hours ago

By region

Amsterdam reachable, 200 ms
Chicago reachable, 189 ms
Paris reachable, 137 ms
San Jose reachable, 131 ms
Singapore reachable, 345 ms
Stockholm reachable, 187 ms
Sydney reachable, 288 ms
Toronto reachable, 222 ms

Timing by region

Connection timing for each region that loaded the page
Region	DNS	TCP	TLS	TTFB
Amsterdam	55 ms	5 ms	18 ms	159 ms
Chicago	114 ms	1 ms	22 ms	178 ms
Paris	43 ms	9 ms	15 ms	108 ms
San Jose	53 ms	3 ms	6 ms	125 ms
Singapore	300 ms	1 ms	7 ms	341 ms
Stockholm	118 ms	1 ms	10 ms	182 ms
Sydney	49 ms	0 ms	6 ms	239 ms
Toronto	105 ms	14 ms	20 ms	178 ms

We check that assemblyai.com loads, not every feature behind it. Test it yourself

History

Incidents AssemblyAI reported

Since Oct 22, 13 days · from status.assemblyai.com

Incidents: 2
Major or critical: 0
Median to resolve: 59m

Elevated errors and delays for Universal-3 Pro, Universal-3.5 Pro and Universal-3.6 Pro
minor · Nov 2
lasted 1h 15m
Real-time STT: 502 errors related to change to temporary tokens
minor · Oct 28
lasted 42m

AssemblyAI website availability

Measured by FlareWarden · assemblyai.com

Last 7 days: 100%
Last 30 days: 100%
Last 90 days: 100%
Avg response, 30 days: 217ms

Last 30 days

18-day clean streak

Oct 5 Today

No downtime Partial Downtime Not measurable No data

Measured by FlareWarden's own checks of assemblyai.com from multiple regions every 6 hours. AssemblyAI's status page sets the verdict above when it reports trouble, but never these figures. Percentages count only checks that returned an availability answer — 18 days measured so far. A dash means that window does not yet hold enough measured days to publish a figure.

What AssemblyAI does

AssemblyAI is a speech AI API. Developers submit recorded audio to its asynchronous /v2/transcript endpoint, stream live audio over WebSockets for real-time transcripts, and send transcripts through LLM Gateway for meeting notes and summaries. Async requests go to api.assemblyai.com in the US by default, while api.eu.assemblyai.com keeps audio and transcripts in the EU. Streaming has its own US and EU data zone endpoints.

What an outage looks like

Async transcripts sit in processing far longer than usual, or end with status error and the message Internal server error. Please retry your request. Webhook deliveries arrive late, sometimes on one endpoint only. Streaming sessions close with code 3005 Session Cancelled, or connect but return no transcript. Incidents often hit one model on one endpoint, such as Universal-3.5 Pro on the EU async endpoint, while the other region works.

What to do about it

Check status.assemblyai.com, which has separate rows for the Asynchronous API, Streaming API, LLM Gateway, Dashboard and Billing, and whose incidents name the endpoint and model affected. Resubmit async jobs that failed with a server error; AssemblyAI can also retry those up to three times, a setting in your account. Retry streams that closed with 1011, as AssemblyAI advises. If an error persists, email support@assemblyai.com with the transcript or session ID.

Is it down for everyone, or just you?

If this page says AssemblyAI is up but it is not loading for you, the problem is between you and them. Run a check against any URL from up to 17 cities across three regions to find out where it breaks.

Test it yourself

Related services

Deepgram

ElevenLabs

AWS

LiveKit

AssemblyAI outage FAQ

Is AssemblyAI down for everyone or just me?

Check status.assemblyai.com first; its incident posts say which endpoint, model and region are affected, so you can match them to your own requests. If it is green, read the error field on the failed transcript or the Error message before a stream closed, which names causes such as an invalid key or low balance. The checks on this page probe www.assemblyai.com, the company website, so an API fault can happen while they stay green.

Will AssemblyAI retry my webhook if my server is down?

Yes. AssemblyAI needs a 2xx answer within 10 seconds, and if it gets none it retries the delivery up to 10 times, 10 seconds apart, before treating it as permanently failed. A 4xx answer stops the retries at once. The payload carries only the transcript ID and status, so you can always fetch the result later with GET /v2/transcript, which also reports the webhook_status_code it received from you.

Does an AssemblyAI EU outage affect the US endpoint?

Usually not. Recent incidents named one endpoint and said the other was unaffected, such as EU async delays in October 2026 while the US endpoint worked. Switching endpoints is not a neutral fix, though: api.eu.assemblyai.com exists to keep audio and transcripts inside the EU, and api.assemblyai.com processes them in the US. Check your data-residency obligations before moving traffic during an incident.

Why am I getting 1008 errors on AssemblyAI streaming?

Usually because of your account, not an outage. Code 1008 covers too many new sessions per minute, a missing or invalid API key, an expired temporary token, a disabled key and insufficient funds. AssemblyAI says to retry with backoff only when the error text reads Too many concurrent sessions; any other 1008 fails again until you fix the key, token or balance. Codes 3005 and 1011 are the server-side ones.

How long do AssemblyAI outages usually last?

Most are resolved within a few hours. Of the 50 most recent incidents on AssemblyAI's status page, the median was resolved in about 1.3 hours, and 40 were closed within four hours. The longest in that set, failing streaming sessions in September 2025, ran just under 12 hours. Backlog incidents can outlast the fix, because queued jobs and webhooks keep arriving late until they drain.

How we measure this

We request AssemblyAI's public endpoint every 6 hours from Fly.io regions across six continents — the most recent check ran from 8 of them. When you open this page and our last check is more than 30 minutes old, we check again on the spot; those checks update the verdict, never the availability figures.
A region counts as down only when it gets no usable HTTP response. A 403 or 429 means the origin answered and refused us, which we report as blocked, never as an outage.
A single failing region is treated as probe noise. Our checks only change their verdict when two consecutive cycles agree.
Response times average only the regions that actually served the page, so a timeout never inflates the number.
Where AssemblyAI publishes an official status feed we read it too, and the verdict is the worse of what it reports and what our checks see: a major outage of AssemblyAI itself is Down, trouble with one part of it is Degraded, problems at a few of a provider's many locations change nothing, and a security notice that marks nothing down is shown but changes nothing either. We check the feed every 3 minutes and read it again within minutes of any change to its incidents or components; a reading counts for 45 minutes after the last check that found nothing changed. The availability figures above are our measurement alone.

Get alerted when AssemblyAI goes down

We check AssemblyAI every 6 hours, and again when you open this page if that was more than 30 minutes ago. Your own monitors run as often as every 30 seconds, from up to 17 cities across three regions, and tell you the moment something breaks.

Start Free Monitoring

Free plan available No credit card required