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: emnify
providerId: emnify
created: '2026-05-25'
modified: '2026-05-25'
reconciled: true
tags:
  - IoT
  - Rate Limiting
  - Quotas
description: Reconciled rate limits for the emnify REST API. Per-IP burst protection, per-account monthly quota, and per-endpoint sustained/burst limits enforced via 429 responses.
sources:
  - https://docs.emnify.com/developers/api-guidelines/rate-limits
  - https://docs.emnify.com/developers/api-guidelines/errors
responseCodes:
  throttled: 429
  quotaExceeded: 429
algorithm: leaky-bucket
quotas:
  - scope: per-account
    metric: requests
    limit: 10000000
    window: month
    description: 10 million API requests per organization per month; resets the first day of each month.
  - scope: per-ip
    metric: requests
    limit: 2000
    window: 5m
    description: 2,000 requests per IP address in any rolling 5-minute window.
limits:
  - scope: endpoint
    pattern: '/api/v1/authenticate'
    rps: null
    window: 5m
    limit: 100
    description: Authentication endpoint allows 100 calls per IP per 5-minute window; cache JWTs aggressively.
  - scope: endpoint-class
    pattern: 'stats and most list endpoints'
    rps: 10
    burst: 5
    description: Standard listing and stats endpoints throttled at 10 req/s sustained with a 5-request burst.
  - scope: endpoint-class
    pattern: 'enhanced endpoints (endpoint stats, SIM retrieval, service/tariff profiles)'
    rps: 100
    burst: 50
    description: Frequently used read endpoints raised to 100 req/s sustained with 50-request burst.
  - scope: endpoint-class
    pattern: 'individual endpoint operations'
    rps: 50
    burst: 25
    description: Mutating endpoint operations capped at 50 req/s with 25-request burst.
recommendations:
  - Cache JWTs and refresh proactively before expiry rather than re-authenticating per call.
  - Use exponential backoff with jitter on 429 responses.
  - For high-volume event consumption, configure the Data Streamer to push events to S3/Kinesis instead of polling /api/v1/event.
  - Use bulk endpoints (/api/v2/endpoint/multi) for fleet operations to stay under per-second limits.