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: FireHydrant
providerId: firehydrant
generated: '2026-08-29'
method: searched
source: https://docs.firehydrant.com/reference/firehydrant-api
created: '2026-05-04'
modified: '2026-08-29'
tags:
- AIOps
- Incident Management
- Rate Limiting
description: >-
  FireHydrant publishes a single, flat API rate limit that applies per ACCOUNT, not per key —
  50 requests every 10 seconds (300 per minute), shared across every API key the organization
  has issued. It is not tiered by plan. This artifact replaces the 2026-05-04 scaffold, whose
  per-tier free/professional/enterprise numbers were invented and are not what FireHydrant
  publishes.
limit_count: 1
headers:
  limit: RateLimit-Limit
  remaining: null
  reset: null
  retryAfter: Retry-After
  policy: null
  note: >-
    FireHydrant documents exactly two response headers on a throttled response — RateLimit-Limit
    (the maximum number of requests in the pool) and Retry-After (seconds to wait). No
    X-RateLimit-Remaining or reset header is documented.
responseCodes:
  throttled: 429
errorBody: '{"error": "rate limit exceeded"}'
limits:
- name: Account API rate limit
  scope: account
  applies_to: all API keys belonging to the account, collectively
  metric: requests
  limit: 50
  window: 10 seconds
  timeFrame: 10s
  equivalent_per_minute: 300
  burst: null
  effective_since: '2023-02-07'
  tier: all
  source: https://docs.firehydrant.com/reference/firehydrant-api
  applies:
  - FireHydrant API
policies:
- name: Retry-After
  description: >-
    On a 429, honor the Retry-After header, which gives the number of seconds to wait before
    retrying. FireHydrant does not document a remaining/reset header, so a client cannot pace
    itself proactively — it must react to the 429.
- name: Shared pool
  description: >-
    Adding API keys does not add capacity. Every key on an account draws from the same 50/10s
    pool, so a noisy automation starves every other integration on the org.
- name: Read replica
  description: >-
    https://api-read.firehydrant.io/v1 is a documented read-only host with a much longer timeout
    for complex reads (may lag the primary by up to 30s). The docs do not state whether it has a
    separate rate-limit pool.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com