specification: API Commons Plans
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/Plans
provider: Inferless
providerId: inferless
created: '2026-06-20'
modified: '2026-06-20'
reconciled: true
tags:
- AI
- ML Inference
- Serverless GPU
- Model Deployment
- Inference
- Plans
description: >-
  Inferless uses usage-based, per-second GPU compute billing. You pay only for
  the inference seconds consumed while a request is being served; when minimum
  replicas are set to zero and there is no active traffic, no charge accrues.
  Rates vary by GPU machine type (T4, A10, A100) and by whether the instance is
  shared (fractional) or dedicated. New accounts receive free credit to start.
notes: >-
  Representative per-second / per-hour GPU rates as published on the Inferless
  pricing page; verify the current machine catalog and rates on the page during
  reconciliation.
sources:
- https://www.inferless.com/pricing
- https://docs.inferless.com
plans:
- id: inferless-free-credit
  name: Free Credit
  type: free
  description: >-
    New users start with free GPU credit and no credit card required, to
    deploy and test models before paying for usage.
  entries:
  - label: Free GPU Credit
    name: free_credit
    type: free
    metric: hours
    limit: 10
    timeFrame: onboarding
    geo: global
    unit: 1
    price: 10 hours free credit (plus promotional $30 credit), no credit card required
    userMultiplied: false
  elements:
  - name: Inference Endpoints
  - name: Model Import
- id: inferless-pay-as-you-go-shared
  name: Pay-as-you-go (Shared Instances)
  type: usage
  description: >-
    Per-second GPU billing on shared (fractional) instances - GPU resources are
    allocated among several users for cost-effective, variable-performance
    serving suited to smaller or infrequent workloads.
  entries:
  - label: T4 (Shared)
    name: t4_shared
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000092/sec ($0.33/hr)
    userMultiplied: false
  - label: A10 (Shared)
    name: a10_shared
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000170/sec ($0.61/hr)
    userMultiplied: false
  - label: A100 (Shared)
    name: a100_shared
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000745/sec ($2.68/hr)
    userMultiplied: false
  elements:
  - name: Inference Endpoints
  - name: Autoscaling to Zero
- id: inferless-pay-as-you-go-dedicated
  name: Pay-as-you-go (Dedicated Instances)
  type: usage
  description: >-
    Per-second GPU billing on dedicated instances reserving the full GPU for a
    single workload for consistent performance.
  entries:
  - label: T4 (Dedicated)
    name: t4_dedicated
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000185/sec ($0.66/hr)
    userMultiplied: false
  - label: A10 (Dedicated)
    name: a10_dedicated
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.000341/sec ($1.22/hr)
    userMultiplied: false
  - label: A100 (Dedicated)
    name: a100_dedicated
    type: usage
    metric: gpu_seconds
    limit: -1
    timeFrame: usage
    geo: global
    unit: 1
    price: $0.001491/sec ($5.36/hr)
    userMultiplied: false
  elements:
  - name: Inference Endpoints
  - name: Dedicated GPU
- id: inferless-enterprise
  name: Enterprise
  type: enterprise
  description: >-
    Volume plan for high-throughput production workloads with discounted
    per-second rates, higher GPU concurrency, extended log retention, custom
    credits, and negotiated terms. Contact Inferless sales.
  entries:
  - label: Enterprise Agreement
    name: enterprise
    type: flat
    metric: contract
    limit: -1
    timeFrame: month
    geo: global
    unit: 1
    price: discounted price (min 100,000 inference requests/month; GPU concurrency 50; 365-day log retention)
    userMultiplied: false
  elements:
  - name: Discounted Per-Second Rates
  - name: Higher GPU Concurrency
  - name: Extended Log Retention
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com