specification: FinOps Framework
specificationVersion: '1.0'
schema: https://www.finops.org/framework/
provider: Refuel
providerId: refuel-ai
created: '2026-06-21'
modified: '2026-06-21'
reconciled: false
tags:
- AI
- LLM
- Data Labeling
- Data Enrichment
- Autolabel
- FinOps
- Cost Management
- FOCUS
description: >-
  FinOps view of Refuel spend. Refuel Cloud is a sales-led enterprise platform
  with no published list pricing; cost is driven by data volume and the number
  of rows transformed through datasets, tasks, and deployed applications, plus
  any custom finetuning. The open-source autolabel library has no Refuel cost -
  spend flows entirely to the underlying LLM provider's per-token rates. Specific
  Refuel Cloud rates are not reconciled in this artifact.
notes: >-
  Refuel does not publish public rates; verify cost structure with Refuel sales
  during reconciliation. For autolabel OSS, track spend at the configured LLM
  provider (OpenAI, Anthropic, Google, HuggingFace, vLLM, or Refuel-hosted).
sources:
- https://www.refuel.ai
- https://docs.refuel.ai
- https://github.com/refuel-ai/autolabel
- 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: Refuel
serviceCategory: AI and Machine Learning
billingModel:
  pricingCategory: Usage-Based
  billingFrequency: Monthly
  billingCurrency: USD
  chargeCategories:
  - Usage
  - Purchase
  - Adjustment
focusColumns:
  ServiceName: Refuel Cloud
  ServiceCategory: AI and Machine Learning
  ProviderName: Refuel
  PublisherName: Refuel
  InvoiceIssuerName: Refuel
  BillingCurrency: USD
  ChargeCategory: Usage
  PricingCategory: Usage-Based
meters:
- name: rows_transformed
  description: Rows of data labeled / enriched through datasets, tasks, and deployed applications.
  unit: rows
  aggregation: sum
  dimensions:
  - account
  - application
  - task
- name: label_requests
  description: Realtime calls to the application label endpoint.
  unit: requests
  aggregation: sum
  dimensions:
  - account
  - application
- name: finetuning_runs
  description: Custom model finetuning runs executed on the platform.
  unit: runs
  aggregation: sum
  dimensions:
  - account
  - model
- name: llm_provider_tokens
  description: >-
    Tokens consumed at the underlying LLM provider when using the open-source
    autolabel library; billed by that provider, not by Refuel.
  unit: tokens
  aggregation: sum
  dimensions:
  - account
  - provider
  - model
principles:
- name: Visibility
  description: Track rows transformed and label requests per application; for autolabel OSS, monitor token burn at the LLM provider.
- name: Allocation
  description: Tag applications, tasks, and API keys per workload/team; map to internal cost centers.
- name: Optimization
  description: Use batch processing and scheduled runs for non-realtime work; prefer smaller / finetuned models where accuracy allows; cache and reuse labeled datasets.
- name: Accountability
  description: Assign owners per project/application; review transformation volume and provider token spend against budget.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com