API reference

The Nexus Media AI REST API lives under /api/v1. Endpoints marked planned are part of the public specification but not yet available. Nothing here is documented unless its contract is final.

Authentication

Two methods are supported: a user bearer token (browser sessions) or a workspace API key for server-to-server calls. API keys use the nma_ prefix and are shown once at creation.

curl https://your-deployment/api/v1/media \
  -H "Authorization: Bearer nma_live_..."

Endpoints

GET/api/v1/healthService health and mode.
GET/api/v1/mediaList media items (cursor pagination).
GET/api/v1/media/{id}Get one media item with transcript metadata.
GET/api/v1/jobsList processing jobs and their status.
POST/api/v1/mediaCreate a media item and request an upload URL.planned
POST/api/v1/video/importImport a supported video URL.planned
POST/api/v1/media/{id}/summariesGenerate a summary from a template.planned
POST/api/v1/media/{id}/chatAsk a source-grounded question.planned

Errors

Every error uses a consistent envelope with a machine-readable code and a request identifier:

{
  "error": {
    "code": "NOT_FOUND",
    "message": "Media item not found.",
    "request_id": "req_8f3k2m",
    "retryable": false
  }
}