generated: '2026-08-17'
method: searched
source: https://gpu-instances.shadow.tech/en/
currency: EUR
plan_count: 3
note: >-
  Shadow GPU publishes its commercial model on its own landing page: three billing models (Pay as you
  go, Monthly plan, Enterprise) crossed with three availability models (Spot, On Demand, Dedicated
  24/7), plus per-configuration starting prices. Values below are quoted verbatim from the page.
  Shadow's per-minute Spot rate is exposed only through an interactive cost simulator whose figure
  changes with configuration, so it is NOT recorded as a fixed price here — only the stated hourly
  starting prices are, because those are published as text.
plans:
- name: Pay as you go
  label: Instant
  price: null
  price_model: usage-based
  billing_period: per-minute
  quota: null
  overage: 'n/a — no commitment'
  included:
  - Instant start with no commitment
  - Ultra-fine per-minute billing
  - Standard weekday support
  limits:
  - 'Instance limit applies; "The limit can be revised after several regular billing cycles. Contact
    our Sales team for quick validation and to avoid service interruption."'
  self_serve: false
  note: 'Billing detail as published: "Pay-as-you-go: per-minute billing, with a detailed summary of
    consumed resources."'
- name: Monthly plan
  label: Predictable
  price: null
  price_model: fixed-monthly
  billing_period: monthly
  quota: 'planned quota (amount not published)'
  overage: 'Additional billing in case of exceeding the planned quota'
  included:
  - Fixed monthly billing
  - Guaranteed 24/7 access to your instances
  - Fixed and predictable monthly budget
  - Success Manager on demand
  self_serve: false
- name: Enterprise
  label: Custom offer
  price: null
  price_model: negotiated
  billing_period: negotiated
  quota: 'Quotas and SLAs negotiated to your activity'
  included:
  - Customized terms
  - Custom integrations (SSO, reporting, connectors)
  - Dedicated Success Manager
  self_serve: false
  contact: 'Talk to an expert (contact form)'
configurations:
- name: RTX 2000 Ada GPU Instance
  gpu: NVIDIA RTX 2000 Ada
  starting_price: '€0.29/h'
  approx_monthly: '≈ €220/month'
  published_specs: 'RT Core 27.7 TFLOPS, Tensor 191.9 TFLOPS'
- name: RTX A4500 GPU Instance
  gpu: NVIDIA RTX A4500
  starting_price: '€0.35/h'
  approx_monthly: '≈ €250/month'
  published_specs: 'RT Core 46.2 TFLOPS, Tensor 189.2 TFLOPS, up to 8 cards in one instance, 20 GB
    VRAM'
availability_models:
- name: Spot
  guarantee: 'Not guaranteed — preemptible based on availability'
  positioning: Performance at best price
  use_cases: [R&D and experimentation, batch processing, CI/CD, one-off calculations]
- name: On Demand
  guarantee: 'Guaranteed — once allocated, availability assured'
  positioning: Flexibility and continuity
  use_cases: [3D and video rendering, AI inference, development, interactive workloads]
- name: Dedicated 24/7
  guarantee: 'Complete guarantee for the entire reservation period'
  positioning: Permanently guaranteed capacity, reserved and isolated
  use_cases: [model training, production AI pipelines, permanent workloads, critical environments]
metering:
  service: CloudKitty (OpenStack rating and usage engine)
  dashboards: 'Real-time cost dashboards (marketing claim)'
  simulator: 'Interactive cost simulator on https://gpu-instances.shadow.tech/en/ comparing PAYG vs
    subscription by configuration, instance count, rate type and daily usage'
  cost_control_note: >-
    Shadow documents that only shelving an instance releases compute resources and stops charges;
    stop, power off, pause and suspend all continue to reserve resources and incur charges. A shelved
    instance is billed only for its stored image.
  api_access_to_billing: 'CloudKitty is listed in the architecture, so usage/rating data is reachable
    through the OpenStack rating API for an authenticated project; no billing API is separately
    documented.'
access:
  self_service_signup: false
  gate: '"Request access" / "Talk to an expert" contact form; the page advertises "⚡ 24h activation"'
  note: >-
    There is no self-service credit-card signup for Shadow GPU. An account is provisioned by Shadow,
    after which the customer's project admin invites further users. This gates the API behind a sales
    step even though the pricing itself is public.
other_products_not_scored_here:
  note: >-
    Shadow's consumer and prosumer subscriptions (Shadow PC, Shadow PC Lite, Shadow PC Pro, Shadow
    Drive) are published at https://shadow.tech/us/shadowpc/offers/ but are end-user desktop
    subscriptions with no API surface, so they are not enumerated as API plans.