generated: '2026-08-17'
method: searched
source: https://www.pyannote.ai/pricing
docs: https://www.pyannote.ai/pricing
currency: EUR
billing_period: monthly
plan_count: 3
model: >-
  Subscription with included monthly credit, then pay-as-you-go at the same per-unit rate. Usage is
  metered in audio-hours for diarization/STT/identification and per-voiceprint for enrollment. Only
  jobs reaching "succeeded" are billed; rejected, failed and canceled jobs are not.
plans:
- name: Developer
  price: 19.00
  currency: EUR
  period: month
  included: 'EUR 19 of usage credit per month, spendable on any endpoint'
  overage: Pay-as-you-go at the same per-unit rates once included credit is consumed
  audience: Solo developers building Voice AI pipelines
  limits:
    workspace_seats: 1
    rate_limit: Default (100 req/min submit, 300 req/min read)
  features:
  - API & Playground access
  - Streaming & Batch diarization
  - Speaker-attributed transcripts
  - Email support
- name: Starter
  price: 99.00
  currency: EUR
  period: month
  recommended: true
  included: 'EUR 99 of usage credit per month, spendable on any endpoint'
  overage: Pay-as-you-go at the same per-unit rates once included credit is consumed
  audience: Small teams integrating speaker AI into workflows
  limits:
    workspace_seats: 3
    rate_limit: Default (100 req/min submit, 300 req/min read)
  features:
  - API & Playground access
  - Streaming & Batch diarization
  - Speaker-attributed transcripts
  - Email support
- name: Enterprise
  price: null
  currency: EUR
  period: null
  pricing_model: Volume-based, contact sales
  audience: Large organizations requiring scale, security, and deployment control
  limits:
    rate_limit: 500 req/min
    concurrency: Higher concurrency, custom limits negotiable
  features:
  - On-premise deployment available
  - Custom limits and higher concurrency
  - Dedicated Slack support
  - Early feature access
unit_rates:
- item: Diarization (Precision-2, batch)
  price: 0.112
  currency: EUR
  unit: audio-hour
- item: STT Orchestration (speech-to-text with diarization)
  price: 0.168
  currency: EUR
  unit: audio-hour
- item: Speaker Identification
  price: 0.015
  currency: EUR
  unit: audio-hour
- item: Voiceprint creation
  price: null
  currency: EUR
  unit: voiceprint
  note: >-
    Billed per voiceprint created rather than by audio duration; the per-unit rate is not published on
    the pricing page. The docs direct customers to the dashboard plan page as the source of truth for
    current per-unit rates.
trial:
  available: true
  length_days: 30
  included: 150 hours of diarization + 10 voiceprints per month
billing_rules:
- rule: minimum charge
  detail: >-
    20-second minimum for every successful /diarize, /identify and /live job. An 8-second job bills as
    20 seconds; a 182-second job bills as 182 seconds.
  source: https://docs.pyannote.ai/administration/billing
- rule: failed jobs are free
  detail: Rejected submissions, failed jobs and canceled jobs are not billed.
- rule: transcription pricing
  detail: >-
    Enabling transcription on /diarize bills the job at the STT Orchestration rate; diarization segments
    are still returned at no additional cost.
- rule: budget cap
  detail: >-
    A monthly budget with email alert thresholds can be set on the billing page; exceeding it causes
    subsequent API requests to be rejected, with a possible enforcement delay.
notes: >-
  Self-serve pricing is fully published with real numbers for two of three tiers, which is unusually
  transparent for an AI infrastructure vendor. The dashboard, not the marketing page, is named as the
  authoritative source for current per-unit rates.