generated: '2026-08-18'
method: searched
source: https://docs.virtuosis.ai/api-reference/recordings/get-recording-analysis
docs:
  - https://docs.virtuosis.ai/api-reference
  - https://docs.virtuosis.ai/api-reference/recordings/get-recording-analysis
  - https://docs.virtuosis.ai/voice-biomarker-api
limit_count: 0
limits: []
guidance:
  - scope: per-recording polling
    operation: getRecordingAnalysis
    rule: Poll every 15 to 30 seconds (recommended).
    floor: Do not poll more frequently than every 5 seconds.
    timeout: Use a five-minute timeout for analysis processing.
    source: https://docs.virtuosis.ai/api-reference
    note: >-
      Stated as client guidance to avoid rate limiting, not as a published quota. No requests-per-window
      number is given.
  - scope: per-request payload
    operation: uploadRecording
    rule: Maximum upload size 50 MB; exceeding it returns 413.
    source: https://docs.virtuosis.ai/api-reference/recordings/upload-recording
    note: A size ceiling, not a rate limit - recorded here because it is the only hard published request bound.
exhaustion:
  status: 429
  operations:
    - uploadRecording
    - getRecordingAnalysis
  description: Rate limit exceeded
  source: openapi/virtuosis-voice-biomarker-api-openapi.yml
  retry_after: not documented
headers:
  published: []
  note: >-
    No rate-limit response headers are documented anywhere on the public surface - no X-RateLimit-*,
    no RateLimit-* (RFC 9331 draft), no Retry-After. The 429 is declared in the OpenAPI but its
    response carries only the standard {error:{type,message}} envelope, so an agent that hits the limit
    is told THAT it was limited but not the window, the remaining budget, or when to retry. Headers
    could not be observed on an authenticated call, but GET /v1.3/ping (the one anonymous operation)
    was probed live on 2026-08-18 and returned HTTP 200 with only content-type, date, cache-control:
    private and request-context - confirming no rate-limit header family is emitted at the edge.
quota_model:
  type: credits
  note: >-
    The commercial constraint is a credit balance, not a request rate. GET /usage/recordings returns
    trial_remaining, trial_granted, included_remaining, included_granted, purchased_remaining and the
    current tier. Uploads consume credits; exhaustion is signalled as 403 UploadLimitExceeded, which is
    distinct from the 429 rate limit.
  buckets:
    - name: trial
      description: One-time trial credits. Consumed before included and purchased credits.
    - name: included
      description: Monthly included credits for the current billing period. Reset each cycle.
    - name: purchased
      description: Purchased credit-pack credits. Roll over until used or refunded.
  source: https://docs.virtuosis.ai/api-reference/usage/get-recordings-usage