generated: '2026-09-19'
method: searched
source: https://airia.ai/docs/build/interface-options/usage_limits
description: >-
  Airia publishes agent-execution rate limits and document-ingestion ceilings per licence tier on its
  Usage Limits page. The OpenAPI declares 429 responses on 10 operations (agent execution, AI-asset
  sync, MCP deployment provisioning and the public governance share endpoints), all returning the
  ProblemDetails shape. No RateLimit-* or X-RateLimit-* response headers are documented, and none
  could be observed without a credential — every unauthenticated request to api.airia.ai returns 401
  before any limit is applied — so the runtime signal an agent would need is not published.
limit_count: 3
limits:
- scope: per-tenant
  tier: Professional
  limit: 3
  unit: agent executions
  window: second
  burst: null
  source: https://airia.ai/docs/build/interface-options/usage_limits
- scope: per-tenant
  tier: Teams
  limit: 4
  unit: agent executions
  window: second
  burst: null
  source: https://airia.ai/docs/build/interface-options/usage_limits
- scope: per-tenant
  tier: Enterprise
  limit: 5
  unit: agent executions
  window: second
  burst: null
  note: >-
    Published as a default that "may be amended based on the Customer's agreement with Airia" — a
    negotiated ceiling, not a fixed one.
  source: https://airia.ai/docs/build/interface-options/usage_limits
quotas:
- scope: per-tenant
  tier: Professional
  limit: 25
  unit: GB
  measure: document ingestion
- scope: per-tenant
  tier: Teams
  limit: 50
  unit: GB
  measure: document ingestion
- scope: per-tenant
  tier: Enterprise
  limit: 100
  unit: GB
  measure: document ingestion
  note: Default, amendable by agreement.
exhaustion:
  status: 429
  body: ProblemDetails (application/json)
  operations_declaring_429: 10
  retry_after: not documented
response_headers:
  documented: []
  observed: []
  note: >-
    No rate-limit headers are documented and none could be observed anonymously; api.airia.ai answers
    401 to every credential-less request. An agent cannot read remaining budget from a response today.
budgets:
  description: >-
    Separate from rate limits, Airia enforces spend budgets at gateway and per-user level (daily,
    weekly, monthly) for model traffic; exhaustion surfaces as 402 on 12 operations.
  source: https://airia.com/ai-platform/cost-optimization