specification: API Commons Plans
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/Plans
provider: Thunder Compute
providerId: thundercompute
created: '2026-06-21'
modified: '2026-06-21'
reconciled: true
tags:
- GPU
- Cloud
- Infrastructure
- AI
- Compute
- Plans
description: >-
  Thunder Compute uses a transparent, usage-based pricing model. On-demand
  virtual GPU instances are billed per minute at a fixed per-GPU-hour rate
  that depends on the GPU model and the instance tier (development /
  prototyping vs. production). The first 4 vCPUs and 100GB of storage are
  included; additional vCPUs, additional persistent storage, and snapshots
  are metered separately. There are no data egress fees. There are no formal
  Free / Developer / Enterprise subscription tiers; users add a payment method
  and pay only for running instances and stored data.
notes: >-
  Per-GPU-hour rates reconciled against the public Thunder Compute pricing
  page on 2026-06-21. GPU availability and rates change; the live
  GET /v1/pricing endpoint and the pricing page are authoritative.
sources:
- https://www.thundercompute.com/pricing
- https://www.thundercompute.com/docs
- https://api.thundercompute.com:8443/v1/pricing
plans:
- id: thundercompute-development
  name: Development (Prototyping) GPU Instances
  type: usage
  description: >-
    On-demand single- or multi-GPU instances for prototyping and
    experimentation, billed per minute at the development per-GPU-hour rate.
  entries:
  - label: RTX A6000 (48GB)
    name: gpu_a6000
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.35 per GPU/hour
    userMultiplied: false
  - label: A100 80GB
    name: gpu_a100_80gb
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.78 per GPU/hour
    userMultiplied: false
  - label: L40 (48GB)
    name: gpu_l40
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.89 per GPU/hour
    userMultiplied: false
  - label: H100 PCIe (80GB)
    name: gpu_h100_pcie
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $1.38 per GPU/hour
    userMultiplied: false
  elements:
  - name: Instances API
  - name: Snapshots API
  - name: tnr CLI
- id: thundercompute-production
  name: Production GPU Instances
  type: usage
  description: >-
    Higher-performance production-tier instances (including NVLink
    configurations) billed per minute at the production per-GPU-hour rate.
  entries:
  - label: L40 (48GB)
    name: gpu_l40_prod
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $1.39 per GPU/hour
    userMultiplied: false
  - label: A100 80GB NVLink
    name: gpu_a100_80gb_nvlink
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $1.79 per GPU/hour
    userMultiplied: false
  - label: H100 PCIe NVLink (80GB)
    name: gpu_h100_pcie_nvlink
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $2.49 per GPU/hour
    userMultiplied: false
  elements:
  - name: Instances API
  - name: Snapshots API
- id: thundercompute-resources
  name: Compute, Storage and Snapshots
  type: usage
  description: >-
    Per-resource add-ons metered alongside GPU instances. First 4 vCPUs and
    100GB of storage are included per instance; usage beyond the included
    amounts is billed separately.
  entries:
  - label: Additional vCPU
    name: extra_vcpu
    type: usage
    metric: vcpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.06 per vCPU/hour (first 4 vCPUs included, 8GB RAM per vCPU)
    userMultiplied: false
  - label: Additional Persistent Storage
    name: extra_storage
    type: usage
    metric: storage_gb_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 100
    price: $0.03 per 100GB/hour (first 100GB included, charged while running)
    userMultiplied: false
  - label: Snapshots
    name: snapshots
    type: usage
    metric: storage_gb_months
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: $0.05 per GB/month
    userMultiplied: false
  - label: Data Egress
    name: egress
    type: usage
    metric: gigabytes
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: $0.00 (no egress fees)
    userMultiplied: false
  elements:
  - name: Per-minute billing
  - name: vCPU
  - name: Storage
  - name: Snapshots
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com