Hatchet · Rate Limits

Hatchet Rate Limits

Rate, throughput, and quota policies for Hatchet Cloud and the open-source Hatchet engine. Cloud throughput limits are sourced from hatchet.run/pricing (Team tier 500 RPS, Enterprise sustained 300M+ runs/month). Engine-level rate-limit primitives are part of the Hatchet API surface itself (the /tenants/{tenant}/rate-limits endpoints) and apply to task execution, not HTTP requests.

Hatchet Rate Limits is the machine-readable rate-limit profile for Hatchet on the APIs.io network, conforming to the API Commons Rate Limits specification.

It captures 6 rate-limit definitions, across the developer, team, scale, enterprise, and self-hosted tiers, measuring requests_per_second and task_runs_per_month.

The profile also includes 6 backoff/retry policies defined and response codes documented for throttled, quotaExceeded, and serviceUnavailable.

Tagged areas include Task Queue, Workflow Engine, Durable Execution, Rate Limiting, and Throughput.

6 Limits Throttle: 429 Quota: 429
Task QueueWorkflow EngineDurable ExecutionRate LimitingThroughput

Limits

Hatchet Cloud Developer Throughput tenant
requests_per_second · second
100
Hatchet Cloud Developer Included Runs tenant
task_runs_per_month · month
100000
Hatchet Cloud Team Throughput tenant
requests_per_second · second
500
Hatchet Cloud Scale Throughput tenant
requests_per_second · second
2000
Hatchet Cloud Enterprise Sustained Capacity contract
task_runs_per_month · month
300000000
Self-Hosted Engine operator-defined
requests_per_second · second
-1

Policies

Engine-Level Rate Limits
Hatchet exposes a first-class rate-limit primitive via the /api/v1/tenants/{tenant}/rate-limits endpoints. Workflows can declare and consume named rate limits that throttle task execution (not HTTP ingress) at configurable units per duration.
Backoff Strategy
Clients should implement exponential backoff with jitter and honor the Retry-After header on 429 and 503 responses.
Burst Handling
Short bursts above the steady-state RPS are tolerated up to the documented burst ceiling before the orchestrator throttles task ingestion.
Tenant Isolation
Quotas and throughput ceilings are scoped per Hatchet tenant; multiple tenants on the same account each receive the full configured allowance independently.
Fair Use (Enterprise)
Enterprise contracts can negotiate sustained capacity well above the 300M-run/month baseline; sustained usage materially in excess of contracted capacity is subject to fair-use review.
Self-Hosting
The open-source Hatchet engine imposes no service-level rate limits; throughput is bounded by the underlying Postgres database, worker pool, and (optionally) RabbitMQ deployment chosen by the operator.