DataCrunch · Rate Limits

Datacrunch Rate Limits

The DataCrunch Public API at https://api.datacrunch.io/v1 is gated by OAuth2 Bearer access tokens scoped to cloud-api-v1. Practical limits are driven by resource quotas (available GPU capacity per location, account spend/credit balance, and per-project instance/volume allowances) rather than published fixed request-per-minute numbers. Serverless container deployments enforce autoscaling bounds via min/max replica counts and queue limits per deployment. Specific numeric API request rate limits are not publicly documented and are not reconciled in this artifact.

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

It captures 5 rate-limit definitions, measuring requests, instances, credit, replicas, and messages.

The profile also includes 3 backoff/retry policies defined and response codes documented for throttled.

Tagged areas include GPU Cloud, Infrastructure, Compute, Inference, and Serverless.

5 Limits Throttle: 429
GPU CloudInfrastructureComputeInferenceServerlessEuropeRate LimitingQuotasThrottling

Limits

API Requests account
requests
see provider documentation
No fixed public RPM/RPD published; requests authorized via OAuth2 Bearer token.
GPU Capacity / Availability account
instances
subject to real-time availability per location
Deploys can be rejected when the requested instance type has no availability.
Account Quota / Balance account
credit
gated by prepaid balance and project quotas
New deployments require sufficient account balance/credit.
Deployment Replicas deployment
replicas
min/max replica count per serverless deployment
Autoscaling bounded by configured min/max replicas and queue length per replica.
Deployment Queue deployment
messages
per-deployment request queue with message TTL
Queue can be purged via the purge-queue endpoint; messages expire per configured TTL.

Policies

OAuth2 Token Scope
Tokens issued via client-credentials are scoped to cloud-api-v1 and restricted to documented Public API endpoints; refresh before expiry.
Capacity-Aware Retry
Treat no-availability responses as capacity signals; retry against alternate instance types or locations rather than tight-looping.
Backoff Strategy
Clients should implement exponential backoff with jitter and honor Retry-After on 429 responses.

Sources