generated: '2026-07-18'
method: searched
source: https://cloud.infini-ai.com/gen-studio/api/get-started/overview.html
provider: Infini-AI (无问芯穹) GenStudio MaaS
summary: >-
  Cross-cutting request/response semantics for the GenStudio large-model service.
  The API is OpenAI-compatible, so most conventions follow the OpenAI Chat
  Completions contract; provider-specific details (base URL, auth, error
  envelope, rate limits) captured from live probes and the docs.
base_url: https://cloud.infini-ai.com/maas/v1
openai_compatible: true
authentication:
  style: bearer-token
  header: 'Authorization: Bearer <API_KEY>'
  ref: authentication/docs-authentication.yml
versioning:
  scheme: uri-path
  current: v1
  path_prefix: /maas/v1
content_type: application/json
streaming:
  supported: true
  style: 'server-sent-events (OpenAI-compatible "stream": true)'
pagination:
  documented: false
  notes: Model-inference endpoints are non-paginated (OpenAI-style single-response / SSE stream).
idempotency:
  supported: false
  notes: No idempotency-key header is documented for the MaaS endpoints.
error_envelope:
  shape: '{"code": <int>, "msg": <string>}'
  ref: errors/docs-problem-types.yml
rate_limiting:
  documented: true
  docs: https://cloud.infini-ai.com/gen-studio/api/usage-and-billing/rate-limit.html
cross_links:
  authentication: authentication/docs-authentication.yml
  errors: errors/docs-problem-types.yml
  lifecycle: lifecycle/docs-lifecycle.yml