Golem · Rate Limits

Golem Cloud Rate Limits

Golem does not publish fixed numeric API rate limits. When self-hosted, the open-source runtime imposes no provider-side rate limits - throughput is bounded only by the infrastructure you run it on and any limits you configure. The managed Golem Cloud hosted service is in Developer Preview and may apply account-level quotas on components, concurrently active workers, and invocation throughput; specific values are not publicly documented and are not reconciled in this artifact.

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

It captures 4 rate-limit definitions, measuring requests, workers, and components.

The profile also includes 2 backoff/retry policies defined and response codes documented for throttled.

Tagged areas include Durable Computing, Serverless, WebAssembly, Workers, and Agents.

4 Limits Throttle: 429
Durable ComputingServerlessWebAssemblyWorkersAgentsRate LimitingQuotasThrottling

Limits

Requests Per Minute (Hosted) account
requests
see provider documentation
Hosted Developer Preview may apply per-account request throttling; values not published.
Concurrently Active Workers account
workers
see provider documentation
Hosted service may cap concurrently active durable workers/agents; self-hosted is unbounded by Golem.
Component Count / Size environment
components
see provider documentation
Hosted service may cap component count or upload size per environment.
Self-Hosted deployment
requests
infrastructure-bound
Open-source self-hosted Golem imposes no provider rate limits; bounded by operator infrastructure.

Policies

Self-Hosted Control
Operators of self-hosted Golem set their own gateway and infrastructure limits; Golem imposes none.
Backoff Strategy
Clients should implement exponential backoff with jitter and honor Retry-After on 429 responses from the hosted service.

Sources