Status

System Duo status

Loading the latest checks.

System One: each model answers a test request every 30 seconds. vllm-sr/auto: a light check every 30 seconds and a real answer every 5 minutes.

Models

OperationalDegradedDownNo data

  1. d3-ultra 27BWaiting for data
    Replicas
    –
    p50
    –
    p90
    –
    Errors, 15 minutes
    –
    Uptime, 24 hours
    –
    24 hours agoNow
  2. d3-flash 9BWaiting for data
    Replicas
    –
    p50
    –
    p90
    –
    Errors, 15 minutes
    –
    Uptime, 24 hours
    –
    24 hours agoNow
  3. vela3-turbo 27BWaiting for data
    Replicas
    –
    p50
    –
    p90
    –
    Errors, 15 minutes
    –
    Uptime, 24 hours
    –
    24 hours agoNow
  4. vela3-flash 9BWaiting for data
    Replicas
    –
    p50
    –
    p90
    –
    Errors, 15 minutes
    –
    Uptime, 24 hours
    –
    24 hours agoNow
  5. vllm-sr/auto Waiting for data
    Pool
    –
    First token p50
    –
    First token p90
    –
    Errors, 15 minutes
    –
    Uptime, 24 hours
    –
    24 hours agoNow

Retired names, now answered by vllm-sr/d3-ultra:

  • d3-turbo
  • d3

Retired names, now answered by vllm-sr/d3-flash:

  • d3-mini
  • d3-nano
  • d3-lite
  • d3-edge

Usage

Requests
–
Tokens: input tokens of the decision models, input and output tokens of vllm-sr/auto
–
Peak requests an hour
–
Users all timeEach API key counts as one user: every key ever created, in any status, including the older sk-pt- keys.
–
Requests
Tokens

Share of requests, last 24 hours

    Server time, last 24 hours p50p90

      Incidents

      Last 3 days