generated: '2026-08-14'
method: searched
source: https://docs.parallel.ai/getting-started/pricing
also:
- https://parallel.ai/pricing
notes: >-
  Parallel does not sell named subscription tiers. It sells one self-serve
  pay-as-you-go account with per-request usage pricing per product, plus signup
  credits and a monthly free allowance. There is therefore ONE plan; the
  commercial surface that matters is the per-product usage rate card below,
  captured verbatim from the published pricing page.
model: pay-as-you-go
currency: USD
plan_count: 1
plans:
- name: Pay as you go
  price: 0
  currency: USD
  billing_period: usage
  self_serve: true
  signup: https://platform.parallel.ai
  included:
    free_requests: Up to 5,000 requests per month for free (per parallel.ai/pricing)
    signup_credit: Up to $80 at signup
    monthly_credit: $5 in free credits per month
  overage: Metered per request at the per-product rates below.
  limits_ref: rate-limits/parallel-rate-limits.yml
enterprise:
  available: true
  note: >-
    Private cloud and on-premises deployments for qualified enterprise customers,
    and higher rate limits on request.
  contact: partnerships@parallel.ai
  support_contact: support@parallel.ai
usage_pricing:
- product: Search API
  unit: per 1,000 requests (default 10 results)
  rates:
  - {tier: turbo, price: 1.00, unit: per 1,000 requests}
  - {tier: basic, price: 5.00, unit: per 1,000 requests}
  - {tier: advanced, price: 5.00, unit: per 1,000 requests}
  - {tier: additional results & excerpts, price: 1.00, unit: per 1,000}
- product: Extract API
  rates:
  - {tier: default, price: 1.00, unit: per 1,000 URLs}
- product: Task API
  unit: per 1,000 Task Runs
  note: >-
    Priced per Task Run (row), not per output field. Failed runs are not billed.
    "-fast" processors carry the same price as their standard counterparts.
  rates:
  - {tier: lite, price: 5.00, unit: per 1,000 runs}
  - {tier: base, price: 10.00, unit: per 1,000 runs}
  - {tier: core, price: 25.00, unit: per 1,000 runs}
  - {tier: core2x, price: 50.00, unit: per 1,000 runs}
  - {tier: pro, price: 100.00, unit: per 1,000 runs}
  - {tier: ultra, price: 300.00, unit: per 1,000 runs}
  - {tier: ultra2x, price: 600.00, unit: per 1,000 runs}
  - {tier: ultra4x, price: 1200.00, unit: per 1,000 runs}
  - {tier: ultra8x, price: 2400.00, unit: per 1,000 runs}
- product: Responses API
  unit: per 1,000 requests
  note: Priced by reasoning.effort. Only successful responses are charged.
  rates:
  - {tier: low, price: 10.00, unit: per 1,000 requests}
  - {tier: medium (default), price: 50.00, unit: per 1,000 requests}
  - {tier: high, price: 250.00, unit: per 1,000 requests}
- product: Monitor API
  unit: per 1,000 executions
  rates:
  - {tier: lite, price: 3.00, unit: per 1,000 executions}
  - {tier: base, price: 10.00, unit: per 1,000 executions}
- product: FindAll API
  unit: fixed cost + per match
  rates:
  - {tier: preview, fixed: 0.10, per_match: 0.00}
  - {tier: base, fixed: 0.25, per_match: 0.03}
  - {tier: core, fixed: 2.00, per_match: 0.15}
  - {tier: pro, fixed: 10.00, per_match: 1.00}
  note: Enrichments add their own per-match cost based on the Task API processor chosen.
- product: Entity Search
  unit: per 1,000 requests (default 100 results)
  rates:
  - {tier: default, price: 5.00, unit: per 1,000 requests}
  - {tier: additional results, price: 0.05, unit: per 1,000}
- product: Chat API
  unit: per 1,000 requests
  rates:
  - {tier: speed, price: 5.00, unit: per 1,000 requests}
  - {tier: lite, price: 5.00, unit: per 1,000 requests}
  - {tier: base, price: 10.00, unit: per 1,000 requests}
  - {tier: core, price: 25.00, unit: per 1,000 requests}
agentic_payments:
  supported: true
  protocols: [Machine Payments Protocol (MPP), x402]
  rails: [Stripe, Tempo stablecoins]
  docs: https://docs.parallel.ai/integrations/agentic-payments