specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Speakeasy
providerId: speakeasy
created: '2026-05-04'
# Provenance stamped 2026-08-11: this artifact was written by the API Evangelist
# bulk sweep dated 2026-05-04, not harvested from the provider. See roadmap#35.
method: generated
modified: '2026-05-05'
reconciled: false
tags:
  - AI
  - Documentation
  - MCP
  - Platform
  - SDKs
  - Rate Limiting
description: Speakeasy does not publish a rate-limits page for its platform APIs.
  The commercial constraint is the per-language operations cap (250 operations per
  language on Free and Business) rather than a per-second throughput limit. MCP /
  AI Control Plane is billed on tool-call requests under enterprise contracts.
sources:
  - https://www.speakeasy.com/pricing
  - https://www.speakeasy.com/docs
notes: No dedicated rate-limit reference page is published; quotas are
  operation-count caps governed by plan, not per-second throughput.
responseCodes:
  throttled: 429
limits:
  - name: Operations per Language (Free / Business)
    scope: account/language
    metric: operations
    limit: 250
    timeFrame: month
    notes: Operations are counted by unique operationId values in the licensed
      OpenAPI spec, per language. Exceeding this cap requires moving to a higher
      tier rather than triggering per-call throttling.
  - name: MCP Tool-Call Requests
    scope: account
    metric: requests
    limit: see enterprise contract
    notes: Speakeasy's AI Control Plane meters MCP tool-call requests under
      enterprise terms; specific per-second limits are not publicly documented.
  - name: Terraform Managed Resources
    scope: account/provider
    metric: resources
    limit: see enterprise contract
    notes: Terraform pricing is bound to the count of managed resources, not API
      throughput.
policies:
  - name: Plan-Based Quota
    description: Speakeasy's enforced limit is the per-language operation count.
      Customers are expected to upgrade tiers rather than throttle on bursts.
  - name: Enterprise Custom Limits
    description: Enterprise contracts may codify per-account throughput limits
      and SLAs that are not part of the public surface.
maintainers:
  - FN: API Evangelist
    email: info@apievangelist.com