specification: FinOps Framework
specificationVersion: '1.0'
schema: https://www.finops.org/framework/
provider: Modal
providerId: modal-labs
created: '2026-07-01'
modified: '2026-07-01'
reconciled: true
tags:
- Serverless
- Compute
- GPU
- AI Infrastructure
- FinOps
- Cost Management
- FOCUS
description: >-
  FinOps view of Modal spend. Modal bills a monthly plan fee (Starter $0,
  Team $250, Enterprise custom) plus per-second usage of CPU cores, memory
  (GiB), and GPU time, and per-GiB-month Volume storage. Free monthly credits
  ($30 Starter, $100 Team) offset compute. Because Modal bills the underlying
  container resources rather than API calls, cost control centers on
  right-sizing GPUs, minimizing container wall-clock time, scaling to zero,
  and choosing cheaper GPU classes when sufficient.
notes: >-
  Per-second rates are subject to change; verify against the Modal pricing
  page during reconciliation. CPU has a 0.125-core minimum; Volumes include
  1 TiB free.
sources:
- https://modal.com/pricing
- https://modal.com/docs/guide/gpu
- 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: Modal
serviceCategory: Compute
billingModel:
  pricingCategory: Usage-Based
  billingFrequency: Monthly
  billingCurrency: USD
  chargeCategories:
  - Usage
  - Purchase
  - Credit
  - Adjustment
focusColumns:
  ServiceName: Modal
  ServiceCategory: Compute
  ProviderName: Modal
  PublisherName: Modal
  InvoiceIssuerName: Modal
  BillingCurrency: USD
  ChargeCategory: Usage
  PricingCategory: Usage-Based
meters:
- name: gpu_seconds
  description: GPU wall-clock time billed per second per GPU class (T4 through B200).
  unit: seconds
  aggregation: sum
  dimensions:
  - workspace
  - gpu_type
  - app
- name: cpu_core_seconds
  description: CPU core-seconds billed at $0.0000131 per core-second (0.125-core minimum).
  unit: core_seconds
  aggregation: sum
  dimensions:
  - workspace
  - app
- name: memory_gib_seconds
  description: Memory billed at $0.00000222 per GiB per second of container runtime.
  unit: gib_seconds
  aggregation: sum
  dimensions:
  - workspace
  - app
- name: volume_gib_month
  description: Volume storage billed at $0.09 per GiB/month with 1 TiB included free.
  unit: gib_month
  aggregation: sum
  dimensions:
  - workspace
  - volume
- name: plan_fee
  description: Monthly subscription plan fee (Starter $0, Team $250, Enterprise custom).
  unit: months
  aggregation: sum
  dimensions:
  - workspace
  - tier
- name: included_credits
  description: Monthly free compute credits applied against usage ($30 Starter, $100 Team).
  unit: usd
  aggregation: sum
  dimensions:
  - workspace
  - tier
principles:
- name: Visibility
  description: Inspect per-app and per-workspace container time and GPU burn in the Modal dashboard and usage exports.
- name: Allocation
  description: Tag apps and functions per workload/team; map workspaces to internal cost centers.
- name: Optimization
  description: >-
    Scale to zero when idle, right-size GPUs (drop from H100 to L4/T4 when
    sufficient), minimize container wall-clock time, use @modal.concurrent to
    raise per-container utilization, and stay under free credit thresholds.
- name: Accountability
  description: Assign owners per app; review per-second compute spend monthly against budget and free-credit allowances.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com