specification: API Commons Plans
specificationVersion: '0.1'
provider: smolagents
providerId: smolagents
name: smolagents Plans and Pricing
description: >
  smolagents is a free open-source library (Apache 2.0). Costs arise when using
  Hugging Face Inference Providers to power agents with cloud-hosted LLMs.
  Hugging Face accounts include monthly inference credits that apply to
  Inference Provider requests routed through Hugging Face. Users who bring
  their own provider API keys (OpenAI, Anthropic, etc.) are billed directly
  by those providers with no Hugging Face charge.
created: '2026-06-12'
modified: '2026-06-12'
reconciled: true
plans:
  - name: Free (Open Source Library)
    description: >
      smolagents library is free to install and use under the Apache 2.0 license.
      A free Hugging Face account is required to access Hugging Face Hub models
      and Inference Providers. Free accounts receive $0.10 in monthly inference
      credits for experimentation.
    type: free
    url: https://huggingface.co/pricing
    quotas:
      - name: Monthly Inference Credits
        description: Included credits toward Hugging Face-routed Inference Provider requests.
        type: metered
        limit: 0.10
        unit: USD
        timeFrame: monthly
      - name: Hub API Rate Limit
        description: Hub API requests per 5-minute window for free authenticated users.
        type: quota
        limit: 1000
        unit: requests
        timeFrame: per-5-minutes
      - name: Resolver Rate Limit
        description: File download (resolver) requests per 5-minute window.
        type: quota
        limit: 5000
        unit: requests
        timeFrame: per-5-minutes
      - name: Pages Rate Limit
        description: Web page requests per 5-minute window.
        type: quota
        limit: 200
        unit: requests
        timeFrame: per-5-minutes
  - name: PRO
    description: >
      Hugging Face PRO account at $9/month provides 20x more inference credits,
      higher Hub API and resolver rate limits, ZeroGPU access, private dataset
      viewer, Spaces Dev Mode, and a PRO badge.
    type: paid
    url: https://huggingface.co/pricing
    price:
      amount: 9.00
      currency: USD
      interval: monthly
    quotas:
      - name: Monthly Inference Credits
        description: Included credits toward Hugging Face-routed Inference Provider requests.
        type: metered
        limit: 2.00
        unit: USD
        timeFrame: monthly
      - name: Hub API Rate Limit
        description: Hub API requests per 5-minute window.
        type: quota
        limit: 2500
        unit: requests
        timeFrame: per-5-minutes
      - name: Resolver Rate Limit
        description: File download requests per 5-minute window.
        type: quota
        limit: 12000
        unit: requests
        timeFrame: per-5-minutes
      - name: Pages Rate Limit
        description: Web page requests per 5-minute window.
        type: quota
        limit: 400
        unit: requests
        timeFrame: per-5-minutes
      - name: ZeroGPU Quota
        description: 8x ZeroGPU quota with highest queue priority.
        type: boolean
        value: true
  - name: Team
    description: >
      Hugging Face Team plan at $20/user/month adds SSO (SAML & OIDC), Storage
      Regions, Audit Logs, Resource Groups, centralized token control, and
      $2.00 per-seat monthly inference credits pooled for the organization.
    type: paid
    url: https://huggingface.co/pricing
    price:
      amount: 20.00
      currency: USD
      interval: monthly
      unit: per-user
    quotas:
      - name: Monthly Inference Credits Per Seat
        description: Inference credits per user seat, pooled across the organization.
        type: metered
        limit: 2.00
        unit: USD
        timeFrame: monthly
      - name: Hub API Rate Limit
        description: Hub API requests per 5-minute window per member.
        type: quota
        limit: 3000
        unit: requests
        timeFrame: per-5-minutes
      - name: Resolver Rate Limit
        description: File download requests per 5-minute window per member.
        type: quota
        limit: 20000
        unit: requests
        timeFrame: per-5-minutes
      - name: Pages Rate Limit
        description: Web page requests per 5-minute window per member.
        type: quota
        limit: 400
        unit: requests
        timeFrame: per-5-minutes
  - name: Enterprise
    description: >
      Hugging Face Enterprise at $50/user/month provides highest storage,
      bandwidth, and API rate limits, SCIM provisioning, advanced security,
      managed billing, legal and compliance support, and dedicated support.
    type: enterprise
    url: https://huggingface.co/pricing
    price:
      amount: 50.00
      currency: USD
      interval: monthly
      unit: per-user
    quotas:
      - name: Monthly Inference Credits Per Seat
        description: Inference credits per user seat, pooled across the organization.
        type: metered
        limit: 2.00
        unit: USD
        timeFrame: monthly
      - name: Hub API Rate Limit
        description: Hub API requests per 5-minute window per member.
        type: quota
        limit: 6000
        unit: requests
        timeFrame: per-5-minutes
      - name: Resolver Rate Limit
        description: File download requests per 5-minute window per member.
        type: quota
        limit: 50000
        unit: requests
        timeFrame: per-5-minutes
      - name: Pages Rate Limit
        description: Web page requests per 5-minute window per member.
        type: quota
        limit: 600
        unit: requests
        timeFrame: per-5-minutes
      - name: Spending Controls
        description: Administrators can set spending limits and disable specific Inference Providers.
        type: boolean
        value: true
  - name: Pay-as-you-Go Inference
    description: >
      After exhausting monthly credits, all account tiers can purchase additional
      inference credits. Hugging Face passes through provider costs at no markup.
      Costs are compute-time based (e.g., a 10-second GPU request at $0.00012/sec
      = $0.0012 per request).
    type: freemium
    url: https://huggingface.co/docs/api-inference/en/pricing
    quotas:
      - name: HF-Inference Compute Billing
        description: >
          Billed per inference request based on compute time multiplied by
          hardware price (e.g., GPU at $0.00012/second).
        type: metered
        unit: USD-per-compute-second