specification: FinOps Framework
specificationVersion: '1.0'
schema: https://www.finops.org/framework/
provider: Cleanlab
providerId: cleanlab
created: '2026-06-21'
modified: '2026-06-21'
reconciled: false
tags:
- AI
- LLM
- Trustworthiness
- Data Quality
- Guardrails
- FinOps
- Cost Management
- FOCUS
description: >-
  FinOps view of Cleanlab spend. The open-source cleanlab library is free and
  incurs no platform cost. Hosted spend comes from the Trustworthy Language
  Model (metered by queries/tokens against the wrapped base LLM), Codex
  validation (metered per validated response), and Cleanlab Studio (data
  curation, AutoML training, and deployed-model inference). Public per-unit
  rates are not consistently published, so cost meters here are unreconciled.
notes: >-
  Per-unit rates are not fully published; verify against Cleanlab pricing and
  billing exports during reconciliation. Open-source cleanlab is free.
sources:
- https://cleanlab.ai/pricing/
- https://help.cleanlab.ai/tlm/
- https://help.cleanlab.ai/studio/
- https://help.cleanlab.ai/codex/
- https://focus.finops.org/focus-specification/v1-3/
alignedWith:
  framework: FinOps Foundation Framework
  frameworkUrl: https://www.finops.org/framework/
  dataSpec: FOCUS
  dataSpecVersion: '1.3'
  dataSpecUrl: https://focus.finops.org/focus-specification/v1-3/
publisherName: Cleanlab
serviceCategory: AI and Machine Learning
billingModel:
  pricingCategory: Usage-Based
  billingFrequency: Monthly
  billingCurrency: USD
  chargeCategories:
  - Usage
  - Purchase
  - Adjustment
focusColumns:
  ServiceName: Cleanlab AI Platform
  ServiceCategory: AI and Machine Learning
  ProviderName: Cleanlab
  PublisherName: Cleanlab
  InvoiceIssuerName: Cleanlab
  BillingCurrency: USD
  ChargeCategory: Usage
  PricingCategory: Usage-Based
meters:
- name: tlm_queries
  description: TLM prompt / trustworthiness-score requests, metered per query.
  unit: queries
  aggregation: sum
  dimensions:
  - account
  - model
  - quality_preset
- name: tlm_tokens
  description: Tokens processed by the wrapped base LLM behind TLM.
  unit: tokens
  aggregation: sum
  dimensions:
  - account
  - model
- name: codex_validations
  description: Responses validated through a Codex project (guardrail + remediation).
  unit: validations
  aggregation: sum
  dimensions:
  - account
  - project
- name: studio_training
  description: AutoML / data-curation compute used to train models in Cleanlab Studio.
  unit: jobs
  aggregation: sum
  dimensions:
  - account
  - project
- name: studio_inference
  description: Predictions served from deployed Cleanlab Studio models via REST.
  unit: predictions
  aggregation: sum
  dimensions:
  - account
  - model
principles:
- name: Visibility
  description: Pull Cleanlab usage/billing exports; track TLM queries, Codex validations, and Studio compute per account and project.
- name: Allocation
  description: Tag API keys and project access keys per workload/team; map to internal cost centers.
- name: Optimization
  description: Use the free open-source cleanlab library where hosted scoring is unnecessary; pick lower-cost base models and quality presets for TLM; batch curation jobs in Studio.
- name: Accountability
  description: Assign owners per project/key; review TLM and platform spend monthly against budget.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com