specification: FinOps Framework
specificationVersion: '1.0'
schema: https://www.finops.org/framework/
provider: Anthropic
providerId: anthropic
created: '2026-05-04'
# Provenance stamped 2026-08-11: this artifact was written by the API Evangelist
# bulk sweep dated 2026-05-04, not harvested from the provider. See roadmap#35.
method: generated
modified: '2026-05-22'
reconciled: true
tags:
  - FinOps
  - Cost Management
  - FOCUS
  - AI
  - Tokens
description: FOCUS-aligned FinOps definition for Anthropic's Claude API. Token-based metering with prompt-caching
  and batch-processing discount lanes.
sources:
  - https://www.anthropic.com/pricing
  - https://platform.claude.com/docs/en/docs/about-claude/pricing
  - https://platform.claude.com/docs/en/api/rate-limits
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: Anthropic
serviceCategory: AI and Machine Learning
billingModel:
  pricingCategory: Usage-Based
  billingFrequency: Monthly
  billingCurrency: USD
  chargeCategories:
    - Usage
    - Purchase
    - Tax
    - Credit
    - Adjustment
  primaryUnit: token
  secondaryUnits:
    - search
    - container_hour
    - session_hour
focusColumns:
  ServiceName: Anthropic API
  ServiceCategory: AI and Machine Learning
  ProviderName: Anthropic
  PublisherName: Anthropic
  InvoiceIssuerName: Anthropic
  PricingCategory: Usage-Based
  PricingUnit: MTok
  BillingCurrency: USD
meters:
  - name: input_tokens
    unit: token
    aggregation: sum
    dimensions:
      - model
      - workspace
      - service_tier
      - inference_geo
  - name: output_tokens
    unit: token
    aggregation: sum
    dimensions:
      - model
      - workspace
      - service_tier
      - inference_geo
  - name: cache_creation_input_tokens
    unit: token
    aggregation: sum
    dimensions:
      - model
      - cache_ttl
  - name: cache_read_input_tokens
    unit: token
    aggregation: sum
    dimensions:
      - model
  - name: web_search_requests
    unit: search
    aggregation: sum
    dimensions:
      - workspace
  - name: code_execution_container_hours
    unit: container_hour
    aggregation: sum
    dimensions:
      - workspace
discountModels:
  - name: Prompt caching
    discount: Cache reads at 0.1x base input price
  - name: Batch API
    discount: 50% off both input and output tokens
  - name: Volume / Enterprise
    discount: Custom; contact sales
unitEconomics:
  - name: Cost per 1M input tokens (Sonnet)
    metric: billed_cost / (input_tokens / 1_000_000)
    target: $3.00
  - name: Cache hit rate
    metric: cache_read_input_tokens / total_input_tokens
    target: '>50%'
  - name: Effective input rate with caching
    metric: (uncached_input * $3 + cached_input * $0.30) / total_input
    target: <$1.50/MTok
principles:
  - name: Visibility
    description: Use the Usage & Cost Admin API to surface per-workspace spend to teams in near real-time.
  - name: Allocation
    description: Tag requests with workspace, application, and feature dimensions for chargeback/showback.
  - name: Optimization
    description: Drive cache hit rates >50%, batch non-interactive workloads (50% discount), pick smallest
      viable model.
  - name: Accountability
    description: Set per-workspace spend limits below tier ceilings to bound team budgets.