Scaleway · Rate Limits
Scaleway Rate Limits
Scaleway exposes a unified REST API across all products under api.scaleway.com, authenticated via the X-Auth-Token header. Per-product quotas are enforced rather than a single global RPS ceiling; each Scaleway product (Instances, Kubernetes, IAM, Object Storage, Generative APIs, etc.) has its own service-specific quotas documented in the developer portal. Public rate-limit values are not centrally published; honor 429 / 503 responses and back off.
2 Limits
Throttle: 429
Cloud ComputingEuropean CloudRate Limiting
Limits
Per-service quotas account/region/service
see service-specific quotas in the Scaleway developer portal
Each Scaleway product (Instance, Kapsule, IAM, Object Storage, Serverless, Generative APIs, etc.) enforces its own quotas and concurrency caps; numbers are not centrally published.
Generative APIs throttling account
model-specific TPS / TPM; see Generative APIs reference
Inference endpoints are throttled per model and per project; quota raises must be requested via support.
Policies
Authentication
Every request must carry an X-Auth-Token header with a valid IAM secret key; permissions are inherited from the IAM principal.
Retry on 429 / 503
Clients should back off on HTTP 429 (throttled) and 503 (service unavailable) using exponential backoff with jitter.
Quota raise via support
Per-resource quotas (instance count, IP count, etc.) can be raised through a Scaleway support ticket.
Regional scoping
Endpoints are zoned (api.scaleway.com/{product}/{version}/zones/{zone}/...), regional, or global; quotas apply per region or zone where applicable.