specification: API Commons Plans
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/Plans
provider: JarvisLabs
providerId: jarvislabs
created: '2026-06-21'
modified: '2026-06-21'
reconciled: true
tags:
- AI
- GPU
- Cloud
- Infrastructure
- Compute
- Plans
description: >-
  JarvisLabs.ai uses a transparent pay-as-you-go model billed per minute of
  compute at published per-GPU-hour rates, with separate per-vCPU-hour rates
  for CPU VMs and a flat per-GB-per-month rate for persistent storage. Pausing
  an instance stops compute billing while retaining storage. There are no
  formal Free / Developer / Enterprise tiers; users recharge a wallet and pay
  for actual usage. The jarvislabs SDK and jl CLI are free.
notes: >-
  Representative on-demand rates as published on the JarvisLabs pricing page;
  GPU availability and rates vary by region and change over time. Storage is
  billed at $0.10/GB per month; egress is free within region.
sources:
- https://jarvislabs.ai/pricing
- https://docs.jarvislabs.ai/
plans:
- id: jarvislabs-pay-as-you-go
  name: Pay-as-you-go
  type: usage
  description: >-
    On-demand GPU and CPU instances billed per minute at published
    per-hour rates, with no monthly minimum.
  entries:
  - label: H100 SXM (80GB)
    name: gpu_h100_sxm
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $2.69 per GPU-hour
    userMultiplied: false
  - label: RTX Pro 6000 Blackwell (96GB)
    name: gpu_rtx_pro_6000
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $1.89 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: $1.49 per GPU-hour
    userMultiplied: false
  - label: A100 (40GB)
    name: gpu_a100_40gb
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.89 per GPU-hour
    userMultiplied: false
  - label: L4 (24GB)
    name: gpu_l4
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.44 per GPU-hour
    userMultiplied: false
  - label: A30 (24GB)
    name: gpu_a30
    type: usage
    metric: gpu_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.41 per GPU-hour
    userMultiplied: false
  - label: Persistent Storage
    name: storage
    type: usage
    metric: gb
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: $0.10 per GB per month
    userMultiplied: false
  - label: Egress (within region)
    name: egress
    type: usage
    metric: gb
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: free within region
    userMultiplied: false
  elements:
  - name: GPU Instances
  - name: CPU VMs
  - name: Templates
  - name: Persistent Filesystems
  - name: Managed Runs
- id: jarvislabs-cpu-vms
  name: CPU VMs
  type: usage
  description: >-
    On-demand CPU-only virtual machines billed per minute at published
    per-hour rates.
  entries:
  - label: Basic (2 vCPU / 8GB)
    name: cpu_basic
    type: usage
    metric: vm_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.05 per hour
    userMultiplied: false
  - label: Standard (4 vCPU / 16GB)
    name: cpu_standard
    type: usage
    metric: vm_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.10 per hour
    userMultiplied: false
  - label: Large (8 vCPU / 32GB)
    name: cpu_large
    type: usage
    metric: vm_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.20 per hour
    userMultiplied: false
  - label: XL (16 vCPU / 64GB)
    name: cpu_xl
    type: usage
    metric: vm_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.40 per hour
    userMultiplied: false
  - label: 2XL (32 vCPU / 128GB)
    name: cpu_2xl
    type: usage
    metric: vm_hours
    limit: -1
    timeFrame: hour
    geo: global
    unit: 1
    price: $0.79 per hour
    userMultiplied: false
  elements:
  - name: CPU VMs
- id: jarvislabs-reserved
  name: Reserved / Committed
  type: enterprise
  description: >-
    Reserved instances and longer-duration (week / month) commitments for
    discounted, dedicated GPU capacity. Contact JarvisLabs for volume terms.
  entries:
  - label: Reserved Capacity
    name: reserved
    type: flat
    metric: contract
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: contact sales
    userMultiplied: false
  elements:
  - name: Reserved GPU Instances
  - name: Committed Duration Discounts
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com