finops: 0.1
focus_alignment: FinOps Foundation FOCUS 1.0
provider: LALAL.AI
provider_id: lalal-ai
updated: '2026-05-25'
billing_model: |
  Subscription + consumable minutes. The license key (X-License-Key) is the FOCUS Resource. Charges accrue from
  subscription plan fees (monthly / annual) and from one-time top-up purchases that grant Fast Queue minutes.
  Minutes are consumed as: file_length_seconds * number_of_selected_stems / 60.

dimensions:
  ServiceCategory: Audio AI / Stem Separation
  ServiceName: LALAL.AI API
  SkuFamily:
    - subscription
    - top-up
    - api-access
  PricingUnit:
    - per-month (subscription)
    - per-year (annual subscription)
    - per-minute-processed (top-up consumption)
  ChargeCategory:
    - Usage      # minutes consumed
    - Purchase   # subscription or top-up purchase
  ChargeFrequency:
    - Recurring (subscription)
    - OneTime (top-up, upgrade)
  ResourceType: license_key
  ResourceId: X-License-Key value
  Region: global (single LALAL.AI endpoint)

cost_drivers:
  - file_length_seconds — primary multiplier for minute consumption.
  - number_of_stems — multistem requests multiply minute consumption per stem.
  - queue_selection — Fast Queue minutes are metered; Relaxed Queue is unlimited on paid plans.
  - top_up_pack — Master ($50 / 750 min), Premium ($190 / 3000 min), Enterprise ($300 / 5000 min).
  - voice_cloning — training and inference may consume additional minutes per LALAL.AI pricing.

unit_economics:
  starter:
    base_cost_usd: 0
    included_minutes: 10
    overage_unit_cost_usd: null
  lite_monthly:
    base_cost_usd: 7.50
    included_fast_minutes: 90
    included_relaxed_minutes: unlimited
  lite_annual:
    base_cost_usd: 90.00
    included_fast_minutes_per_month: 90
  pro_monthly:
    base_cost_usd: 15.00
    included_fast_minutes: 250
    api_access: true
  pro_annual:
    base_cost_usd: 180.00
    included_fast_minutes_per_month: 250
    api_access: true
  topups:
    master:
      cost_usd: 50.00
      minutes: 750
      effective_rate_usd_per_minute: 0.0667
    premium:
      cost_usd: 190.00
      minutes: 3000
      effective_rate_usd_per_minute: 0.0633
    enterprise:
      cost_usd: 300.00
      minutes: 5000
      effective_rate_usd_per_minute: 0.0600

observability:
  remaining_minutes_endpoint: POST /api/v1/limits/minutes_left/
  invoice_history: Available via the LALAL.AI account dashboard.
  webhooks: Not publicly documented at the API v1 level; jobs are polled via /api/v1/check/.
  recommended_tags:
    - cost_center
    - project
    - environment

optimization_levers:
  - Use Relaxed Queue for non-time-sensitive batch jobs to preserve Fast Queue minutes.
  - Prefer multistem requests over multiple single-stem requests when extracting several stems from one source —
    same consumption math, fewer round-trips.
  - Cache stem_track / back_track downloads locally; signed URLs are temporary, but re-running a job re-bills.
  - Negotiate custom enterprise pricing (support@lalal.ai) for sustained high-volume API usage.