specification: FinOps Framework
specificationVersion: '1.0'
schema: https://www.finops.org/framework/
provider: Lightly
providerId: lightly-ai
created: '2026-06-21'
modified: '2026-06-21'
reconciled: false
tags:
- AI
- Computer Vision
- Data Curation
- Active Learning
- Embeddings
- FinOps
- Cost Management
- FOCUS
description: >-
  FinOps view of LightlyOne platform spend. Lightly bills via plan
  subscription (free Community, paid Team, custom Enterprise) rather than a
  granular per-call meter, so cost is driven by the chosen tier and the
  dataset-size / media / collaboration entitlements it unlocks. The main cost
  levers are input dataset volume, LightlyOne Worker compute-run frequency, and
  cloud-storage egress on the customer's own S3 / GCS / Azure datasource (paid
  to the cloud provider, not Lightly). The open-source `lightly` SDK is free.
  Figures are not reconciled against a current rate card.
notes: >-
  Lightly does not publish a public usage-metered rate card; Enterprise is
  quoted by sales. Treat meters below as cost-allocation dimensions for
  internal tracking rather than billed line items, and verify on reconciliation.
sources:
- https://www.lightly.ai/pricing
- https://docs.lightly.ai/
- 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: Lightly
serviceCategory: AI and Machine Learning
billingModel:
  pricingCategory: Subscription-Based
  billingFrequency: Monthly
  billingCurrency: USD
  chargeCategories:
  - Usage
  - Purchase
focusColumns:
  ServiceName: LightlyOne
  ServiceCategory: AI and Machine Learning
  ProviderName: Lightly
  PublisherName: Lightly
  InvoiceIssuerName: Lightly
  BillingCurrency: USD
  ChargeCategory: Usage
  PricingCategory: Subscription-Based
meters:
- name: plan_subscription
  description: Monthly platform subscription for the selected tier (Team / Enterprise).
  unit: subscription
  aggregation: sum
  dimensions:
  - account
  - plan
- name: input_dataset_samples
  description: Number of input samples curated, bounded by the plan dataset-size quota.
  unit: samples
  aggregation: max
  dimensions:
  - account
  - dataset
- name: worker_runs
  description: LightlyOne Worker selection / active-learning compute runs executed.
  unit: runs
  aggregation: sum
  dimensions:
  - account
  - dataset
  - worker
- name: stored_data_bytes
  description: Data and embeddings stored against the platform / datasource.
  unit: bytes
  aggregation: max
  dimensions:
  - account
  - dataset
- name: datasource_egress_bytes
  description: Cloud-storage egress on the customer datasource (billed by the cloud provider, not Lightly).
  unit: bytes
  aggregation: sum
  dimensions:
  - account
  - datasource
principles:
- name: Visibility
  description: Track input dataset size against the plan quota and count worker runs per dataset to understand cost drivers.
- name: Allocation
  description: Tag datasets and worker labels per team / project to map platform spend to cost centers.
- name: Optimization
  description: Right-size selection runs (fewer, well-configured runs over broad re-runs); keep datasource and Lightly compute in the same cloud region to limit egress; use the free OSS SDK for offline experimentation.
- name: Accountability
  description: Assign dataset / project owners and review tier entitlement vs. actual dataset volume before renewal.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com