GET /v1/files/{file_id}/artifacts — Atom Files Developers

← API referenceFile intelligence · API access unavailable

API REFERENCE / FILE INTELLIGENCE

`GET` /v1/files/{file_id}/artifacts

List derived outputs for an authorized file version.

Request fields

Location	Field	Type	Required	Description
Header	`Authorization`	string	Yes	Credential scoped to the authorized resource.
Path	`file_id`	string	Yes	Stable Atom file ID.
Query	`version_id`	string	Recommended	Limit results to an immutable content version.
Query	`kind`	string (extensible)	No	Output kind, such as page_ocr, transcript, thumbnail, or document_summary. Examples · extensible`thumbnailpage_ocrtranscriptdocument_summarystructured_table`
Query	`cursor`	string	No	Opaque pagination cursor bound to the selected version and filters.

Request example

Example request

GET /v1/files/file_789/artifacts?version_id=ver_012&kind=page_ocr
Authorization: Bearer <scoped token>

HTTP status codes

Status	Meaning
`200 OK`	Success response shown below.
`400 Bad Request`	Invalid version, kind, or cursor.
`401 Unauthorized`	Missing, expired, or invalid credential.
`403 Forbidden`	Credential is valid but lacks the required scope or role.
`404 Not Found`	File or version is not visible to this caller.
`429 Too Many Requests`	Rate limit reached; honor Retry-After when provided.
`503 Service Unavailable`	Service temporarily unavailable; retry safely.

Response fields

Field	Type	Example
`file_id`	string	file_789
`version_id`	string	ver_012
`items`	object[]	[1 shown]
`items[].id`	string	art_123
`items[].kind`	string (extensible)	page_ocr Examples · extensible`thumbnailpage_ocrtranscriptcaptions_srtcaptions_vttdocument_summarystructured_table`
`items[].media_type`	string	text/plain
`items[].page_number`	integer	1
`items[].source_version_id`	string	ver_012
`items[].processor_id`	string	atom-ocr-v1
`items[].created_at`	string (date-time)	2026-10-20T18:02:00Z
`next_cursor`	string \| null	null

Response example

Example response

200 OK
Content-Type: application/json

{
  "file_id": "file_789",
  "version_id": "ver_012",
  "items": [
    {
      "id": "art_123",
      "kind": "page_ocr",
      "media_type": "text/plain",
      "page_number": 1,
      "source_version_id": "ver_012",
      "processor_id": "atom-ocr-v1",
      "created_at": "2026-10-20T18:02:00Z"
    }
  ],
  "next_cursor": null
}

How it works

Outputs may include thumbnails, OCR, transcripts, SRT/WebVTT captions, extracted images, page or document summaries, and structured tables. Availability depends on file type and processing policy. Every artifact identifies its source version and processor.