specification: API Commons Plans
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/Plans
provider: Cerebrium
providerId: cerebrium
created: '2026-06-20'
modified: '2026-06-20'
reconciled: true
tags:
- AI
- GPU
- Serverless
- Inference
- ML Infrastructure
- Plans
description: >-
  Cerebrium combines flat monthly plan tiers (Hobby, Standard, Enterprise) with
  usage-based, per-second compute billing. You pay for GPU, CPU, and memory only
  while a function is running, plus persistent storage per GB-month. GPU rates
  are charged per second by accelerator type. The Hobby plan is free plus
  compute; the Standard plan adds a $100/month platform fee plus compute with
  unlimited apps and seats; Enterprise is custom with volume discounts and
  unlimited GPU concurrency.
notes: >-
  Per-second GPU, CPU, memory, and storage rates as published on the Cerebrium
  pricing page; verify current rates and the full GPU catalog on the pricing
  page during reconciliation.
sources:
- https://www.cerebrium.ai/pricing
- https://www.cerebrium.ai/docs
plans:
- id: cerebrium-hobby
  name: Hobby
  type: free
  description: >-
    Free platform plan with basic features and community support; compute is
    billed separately per second.
  entries:
  - label: Platform Fee
    name: hobby_platform_fee
    type: flat
    metric: subscription
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: $0 (compute billed separately)
    userMultiplied: false
  elements:
  - name: Serverless GPU/CPU Deployments
  - name: Community Support
- id: cerebrium-standard
  name: Standard
  type: paid
  description: >-
    Paid platform plan with unlimited apps and seats; compute is billed
    separately per second on top of the monthly platform fee.
  entries:
  - label: Platform Fee
    name: standard_platform_fee
    type: flat
    metric: subscription
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: $100/month + compute
    userMultiplied: false
  elements:
  - name: Unlimited Apps
  - name: Unlimited Seats
- id: cerebrium-compute
  name: Per-Second Compute
  type: usage
  description: >-
    Usage-based compute billed by the second for the duration each function
    runs. GPU rates are per accelerator type; CPU and memory are billed per
    vCPU-second and GB-second.
  entries:
  - label: B200 GPU
    name: gpu_b200
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.00167 per second
    userMultiplied: false
  - label: H200 GPU
    name: gpu_h200
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.001166 per second
    userMultiplied: false
  - label: H100 GPU
    name: gpu_h100
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000944 per second
    userMultiplied: false
  - label: RTX PRO 6000 GPU
    name: gpu_rtx_pro_6000
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000694 per second
    userMultiplied: false
  - label: A100 80GB GPU
    name: gpu_a100_80gb
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000583 per second
    userMultiplied: false
  - label: A100 40GB GPU
    name: gpu_a100_40gb
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000555 per second
    userMultiplied: false
  - label: L40s GPU
    name: gpu_l40s
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000542 per second
    userMultiplied: false
  - label: A10 GPU
    name: gpu_a10
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000306 per second
    userMultiplied: false
  - label: L4 GPU
    name: gpu_l4
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000222 per second
    userMultiplied: false
  - label: T4 GPU
    name: gpu_t4
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000164 per second
    userMultiplied: false
  - label: CPU
    name: cpu
    type: usage
    metric: vcpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.00000655 per vCPU per second
    userMultiplied: false
  - label: Memory
    name: memory
    type: usage
    metric: gb_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.00000222 per GB per second
    userMultiplied: false
  - label: Persistent Storage
    name: storage
    type: usage
    metric: gb_month
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: $0.05 per GB/month (first 100GB free)
    userMultiplied: false
  elements:
  - name: GPU Inference
  - name: CPU Inference
  - name: Persistent Volumes
- id: cerebrium-enterprise
  name: Enterprise
  type: enterprise
  description: >-
    Custom enterprise plan with volume discounts, unlimited GPU concurrency, and
    dedicated support. Contact Cerebrium sales.
  entries:
  - label: Enterprise Agreement
    name: enterprise
    type: flat
    metric: contract
    limit: -1
    timeFrame: year
    geo: global
    unit: 1
    price: contact sales
    userMultiplied: false
  elements:
  - name: Volume Discounts
  - name: Unlimited GPU Concurrency
  - name: Dedicated Support
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com