●Incident
Deepgram outage.
ResolvedDEGRADATION · HIGHOct 16, 2026, 03:03 PM · about 20 hours ago · Duration: 3h 24min
Incident Summary
Incident Summary
API Deepgram is currently DOWN in asia.
Diagnostic Details
Error Type: http HTTP Status: 503 Error Message: API returned HTTP 503 Retry Attempts: 1 (all failed)
Validation Results
Validation Reason: DOWN in 2 regions: us-east, asia Consecutive DOWN Checks: 2 Time Between Checks: 370s
Cross-Region Impact
Also DOWN in: us-east Total Affected Regions: 2 Still UP in: europe, us-west
Severity Assessment
Level: MEDIUM Limited regional impact or recoverable error type.
Detected by Probecast Health Monitor - 2026-10-16T15:03:58.574Z
Timeline
●Started
Oct 16, 2026, 03:03 PM
about 20 hours ago
●Resolved
Oct 16, 2026, 06:28 PM
about 17 hours ago
●Duration
3h 24min
Total downtime
Evidence Logs3 checks
| Time | Region | Status | Latency | Code |
|---|---|---|---|---|
| Oct 16, 03:03 PM | asia | down | 232ms | 503 |
| Oct 16, 03:00 PM | us-east | down | 65ms | 503 |
| Oct 16, 02:57 PM | asia | down | 233ms | 503 |
How many of your clients depend on Deepgram?
When Deepgram slows down, every client that depends on it feels it at once, and the first call usually goes to the agency. Probecast watches your clients' sites and the providers underneath them, shows which clients an outage touches, and turns the month into a report your client can trust.
One provider breaks. See every client it touches.
Ordered by how many clients it hits, so you message the right people first.
Each client's site and store, checked from four regions.
US West & East, Europe and Asia — a regional outage never hides.
A monthly report under your brand, reviewed by your team before it goes out.
Uptime, incidents, traffic, revenue and support — signed off by a person.
Your team scoped per client. Every change on the record.
Freelancers see only their clients; every change lands in an append-only audit log.
Alerts land on email, Slack, push or webhooks — with the client's name on them.