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: Apideck
providerId: apideck
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:
  - Rate Limiting
  - Unified API
  - Integrations
description: Apideck does not publish a single global rate-limit number for its Unify API. Effective limits
  are determined by (1) Apideck's own per-application throttling and (2) the downstream connector's native
  rate limits (Salesforce, QuickBooks, Xero, HubSpot, etc.) which Apideck proxies and surfaces on 429.
  Apideck recommends honoring upstream Retry-After headers and using exponential backoff. Plan tier governs
  Active Consumer count and category access rather than a documented per-second cap.
notes: No public per-second/per-minute numbers are documented. Limits are largely connector-dependent.
sources:
  - https://www.apideck.com/pricing
  - https://developers.apideck.com/apis
responseCodes:
  throttled: 429
  serviceUnavailable: 503
headers:
  retryAfter: Retry-After
limits:
  - name: Apideck platform throttling
    scope: application
    metric: varies
    limit: see provider documentation
    notes: Apideck does not publish a fixed per-second number; throttling is applied at the application
      and consumer level to protect shared infrastructure.
  - name: Connector (downstream) rate limits
    scope: connector
    metric: varies
    limit: governed by downstream connector
    notes: When an upstream SaaS (Salesforce, QuickBooks, Xero, HubSpot, etc.) throttles, Apideck propagates
      a 429 with the upstream Retry-After.
policies:
  - name: Backoff Strategy
    description: Implement exponential backoff with jitter; honor Retry-After when returned by Apideck
      or the underlying connector.
  - name: Connector Awareness
    description: Effective throughput depends on the downstream system's native limits. High-volume use
      cases should partition workload across consumer connections and respect per-connector quotas.
  - name: Fair Use
    description: Sustained traffic that materially impacts shared infrastructure may be throttled regardless
      of plan tier; engage Apideck support to align on burst patterns.
  - name: Webhook Replay
    description: Use webhooks plus subscribe/poll patterns rather than tight read loops to reduce request
      volume against rate-limited connectors.
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com