Voice Stack · platform status

We're experiencing an outage.

Last check Wed, 18 Nov 2026 05:19:56 GMT

Components

latency = live probe from the studio server
Voice API
https://voice-api.xpin.ae — HTTP & WebSocket endpoints
unreachable: HTTP 502
— Outage
Speech engine (GPU)
Voice Stack speech model on the inference node
cannot reach the API
— Unknown
Transcription engine (GPU)
Scribe speech-to-text model on the inference node
unreachable: HTTP 502
— Outage
Web studio
https://voice.xpin.ae — app, login, history
serving · up 7d 0h
— Operational
Object storage
MinIO — generated audio & voice references
bucket reachable
232 ms Operational
Database
Accounts, voices, generation history
query OK
17 ms Operational
Middleware API gateway
https://middleware.xpin.ae — access tokens, quotas and metering for published APIs
0/0 upstreams up · 0 endpoints published · 0 active tokens
23 ms Operational

Machines

voice.xpin.ae
Web studio & app server
Operational
Location
Azure · UAE North · AKS
Spec
Next.js on AKS
Now
this page is served from here
voice-api.xpin.ae
API gateway
Outage
Location
Azure · UAE North · AKS
Spec
nginx + reverse SSH to GPU
Now
no response
GPU inference node
Speech engine
Unknown
Location
India · Delhi
Spec
NVIDIA H200 · cuda · fp16
Now
model not loaded
middleware.xpin.ae
API gateway & monetization
Operational
Location
Azure · UAE North · AKS
Spec
2 vCPU · 3 GB RAM · FastAPI
Now
23 ms round-trip
s3.xpin.ae
Object storage (MinIO)
Operational
Location
Azure · UAE North · AKS
Spec
S3-compatible · bucket voice-stack
Now
232 ms
Recent latency
API health— ms
collecting…
Synthesis (short clip)— ms
collecting…
1 checks kept since the studio server started.
Availability by region
  • AEUnited Arab Emirates Operational
  • INIndia Operational

Voice Stack is offered in the UAE and India. Other regions receive HTTP 451 by design.

Build
Studio
v0.1.0
Server uptime
7d 0h
Runtime
Node v22.23.3
Engine
— · 24000 Hz

Monitors can poll /api/status (HTTP 503 while any component is down).