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: Edge Impulse
providerId: edge-impulse
generated: '2026-09-06'
method: searched
source: https://www.edgeimpulse.com/pricing, https://docs.edgeimpulse.com/apis/studio, https://docs.edgeimpulse.com/apis/ingestion,
  https://studio.edgeimpulse.com/openapi.yml
created: '2026-05-04'
modified: '2026-09-06'
tags:
- Artificial Intelligence
- Embedded Machine Learning
- Edge AI
- Quotas
description: 'Edge Impulse publishes NO request-rate limits and NO rate-limit response headers. No 429
  appears anywhere in the 559-operation published OpenAPI, and no X-RateLimit-*, RateLimit-* or Retry-After
  header is documented on any of the three APIs. This is an honest zero, and it is a real gap for an agent:
  there is no runtime signal telling a caller how close it is to a wall. What Edge Impulse does meter
  is RESOURCE consumption — compute minutes per job, memory, project and collaborator counts, and upload
  size — plus hard payload caps on the Ingestion API. Those are recorded below. This REPLACES a 2026-05-04
  bulk-sweep scaffold that invented per-minute and per-month request quotas across three tiers that Edge
  Impulse does not sell.'
limit_count: 0
quota_count: 8
headers:
  limit: null
  remaining: null
  reset: null
  retryAfter: null
  policy: null
  note: None documented. Verified against the full published OpenAPI and the Studio and Ingestion API
    reference pages on 2026-09-06.
responseCodes:
  throttled: null
  note: No 429 is declared in the published spec. The Studio API declares only 200 (555 operations) and
    302 (4). The Ingestion API documents 200, 400, 401, 421 and 500 — no throttling code among them.
limits: []
quotas:
- tier: developer
  name: Compute time per job
  scope: per-job
  metric: minutes
  limit: 60
  source: https://www.edgeimpulse.com/pricing
  note: Jobs started through the API count against the same budget as jobs started in Studio.
- tier: enterprise
  name: Compute time per job
  scope: per-job
  metric: minutes
  limit: -1
  source: https://www.edgeimpulse.com/pricing
  note: Unlimited.
- tier: developer
  name: CPU compute memory
  scope: per-job
  metric: gigabytes
  limit: 16
  source: https://www.edgeimpulse.com/pricing
- tier: enterprise
  name: CPU compute memory
  scope: per-job
  metric: gigabytes
  limit: null
  source: https://www.edgeimpulse.com/pricing
  note: Configurable.
- tier: developer
  name: Private projects
  scope: per-account
  metric: projects
  limit: 3
  source: https://www.edgeimpulse.com/pricing
- tier: developer
  name: Collaborators per project
  scope: per-project
  metric: users
  limit: 3
  source: https://www.edgeimpulse.com/pricing
- tier: all
  name: Files per ingestion request
  scope: per-request
  metric: files
  limit: 1000
  source: https://docs.edgeimpulse.com/apis/ingestion
- tier: all
  name: Single file size on ingestion
  scope: per-file
  metric: megabytes
  limit: 100
  source: https://docs.edgeimpulse.com/apis/ingestion
limit_increase:
  operationId: requestEnterpriseLimitsIncrease
  path: POST /api/organizations/{organizationId}/limits-request
  increasable:
  - users
  - projects
  - compute
  - storage
  docs: https://docs.edgeimpulse.com/apis/studio/organizations/request-organization-limits-increase
  note: Organization limits are raised by filing a request through the API; there is no self-serve upgrade
    and no published turnaround.
project_overrides:
- operationId: setProjectComputeTimeLimit
  path: POST /api/{projectId}/compute-time-limit
  note: JWT-only and not available to all users.
- operationId: setProjectFileSizeLimit
  note: JWT-only.
  path: POST /api/{projectId}/dsp-file-size-limit
policies:
- name: No published backoff guidance
  description: Edge Impulse does not document a retry or backoff policy. Because the API is not idempotent,
    a naive retry of a job-start POST starts a second job and consumes compute twice.
- name: Quota exhaustion is not signalled at the HTTP layer
  description: Exceeding a compute or project quota surfaces as a failed job or a success:false envelope,
    not as a distinct status code an agent can branch on.
related:
  plans: plans/edge-impulse-plans-pricing.yml
  conventions: conventions/edge-impulse-conventions.yml
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com