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: Red5
providerId: red5
created: '2026-05-04'
modified: '2026-09-04'
generated: '2026-09-04'
method: searched
source:
  - https://www.red5.net/docs/red5-pro/development/api/stream-manager-2.0/
  - https://www.red5.net/red5-cloud-pricing/
  - https://www.red5.net/pricing/
  - openapi/*.yml
reconciled: true
tags:
  - Streaming
  - Real-Time Video
  - Rate Limiting
description: >-
  Red5 publishes no API rate limit of any kind. Re-searched 2026-09-04 across the
  Stream Manager 2.0 API documentation, both pricing pages and all ten OpenAPI
  documents: there is no requests-per-window ceiling, no 429 response declared on any of
  the 26 operations, and no RateLimit-*, X-RateLimit-* or Retry-After header documented
  or returned in any declared response. The 2026-05-04 record said the site was
  unreachable (403); it is reachable now, and the honest answer is that the limits are
  genuinely absent rather than unfetched.
limit_count: 0
responseCodes: {}
limits: []
headers: []
status_on_exhaustion: null
what_bounds_capacity:
  model: deployment-and-quota, not request-rate
  red5_pro: >-
    Self-hosted. Concurrency is bounded by the customer's own node count and licensed
    tier — Developer Pro caps at 100 connections with no clustering and no autoscale;
    Startup, Growth Pro and Enterprise are "unlimited connections" with autoscaling on
    AWS, Google Cloud, Azure or DigitalOcean.
  red5_cloud: >-
    Managed. Bounded by metered plan quotas rather than request rate — 50 GB streaming
    and 6,000 instance hours per month on the free Pay-As-You-Grow tier, 10,000 GB /
    10,000 instance hours on Growth, 20,000 GB / 10,000 instance hours on Enterprise,
    with published per-GB and per-instance-hour overage rates. Exceeding a quota bills
    an overage; it does not return an error status.
  cluster_controls: >-
    Stream Manager 2.0 NodeGroupConfig carries capacity and limit expressions (min() and
    max() added in server 16.0.0) and a max-publishers/max-subscribers example. Those are
    operator-configured admission controls on the customer's own cluster, not a
    provider-imposed API rate limit.
notes: >-
  An agent calling a Red5 deployment gets no runtime backpressure signal from the API.
  There is nothing to read on a response to decide whether to slow down, and no declared
  429 to catch. See conventions/red5-conventions.yml.