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: Windmill
providerId: windmill-dev
created: '2026-07-02'
modified: '2026-07-02'
reconciled: false
tags:
- Developer Platform
- Workflows
- Job Orchestration
- Rate Limiting
- Quotas
description: >-
  Windmill does not publish fixed numeric per-endpoint request-rate limits for
  its REST API. On self-hosted Community Edition, throughput is bounded only by
  your own infrastructure and by how many workers you run - each worker pulls
  and executes jobs from the shared Postgres-backed queue, so overall execution
  concurrency scales with the size of your worker fleet. On Windmill Cloud and
  Enterprise, effective throughput is governed by your provisioned Compute Units
  (worker capacity) and seat entitlements rather than by a per-minute API cap.
  Application-level flow control is expressed through per-script/per-flow
  concurrency limits and named concurrency groups, plus worker group tags that
  route and cap specific workloads.
notes: >-
  Numeric per-account or per-endpoint API limits are not documented as of the
  review date. Control execution load with concurrency limits, concurrency
  groups, worker group sizing, and job timeouts. The Community Edition free tier
  additionally caps SSO users (10), workspaces (3), groups (4), and emails
  (100/day). Verify Cloud/Enterprise allowances on the pricing page.
sources:
- https://www.windmill.dev/docs/core_concepts/worker_groups
- https://www.windmill.dev/docs/core_concepts/concurrency_limits
- https://www.windmill.dev/pricing
- https://github.com/windmill-labs/windmill
responseCodes:
  throttled: 429
limits:
- name: REST API Requests
  scope: account
  metric: requests
  limit: not published
  notes: No fixed numeric request-rate limit is documented for the Windmill REST API.
- name: Self-Hosted Execution Throughput
  scope: deployment
  metric: jobs
  limit: worker-fleet-bound
  notes: Concurrency scales with the number and size of workers you run; bounded by your own hardware.
- name: Compute Units
  scope: account
  metric: worker_capacity
  limit: per plan
  notes: On Cloud/Enterprise, execution capacity is provisioned in Compute Units (about 2 worker-GB-month each).
- name: Per-Script / Per-Flow Concurrency
  scope: runnable
  metric: concurrent_executions
  limit: configurable
  notes: Each script or flow can declare a maximum number of concurrent executions within a time window.
- name: Concurrency Groups
  scope: group
  metric: concurrent_executions
  limit: configurable
  notes: Named concurrency groups cap combined concurrency across multiple runnables that share the group.
- name: Community Free Tier Caps
  scope: instance
  metric: various
  limit: 10 SSO users / 3 workspaces / 4 groups / 100 emails per day
  notes: Structural caps on the free Community Edition, not request-rate limits.
policies:
- name: Queue Backpressure
  description: Jobs are enqueued in a Postgres-backed queue and pulled by available workers; when workers are saturated, jobs wait in the queue rather than being dropped.
- name: Concurrency Limits
  description: Per-runnable and per-group concurrency limits smooth spikes and protect downstream systems from overload.
- name: Retries and Timeouts
  description: Scripts and flow steps support configurable retries with backoff and per-job timeouts to bound runaway executions.
- name: Backoff Strategy
  description: API clients should implement exponential backoff with jitter and honor Retry-After on any 429 response.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com