Create segments¶

Create a segment in the source’s current snapshot. Provide text, OCR text, or an image description for the new segment.

POST https://api.moi.matrixorigin.cn/v5/semantic-models/{model_id}/sources/{source_row_id}/segments

Preparation before calling¶

First read the document and obtain data.etag as the write precondition. Prepare a personal access token with access to the target workspace and the workspace ID.

The example below uses:

$AI_STUDIO_API_KEY: The actual personal access token, passed through the X-API-Key Header.
$WORKSPACE_ID: Target workspace ID, passed through X-Workspace-ID Header.
$MODEL_ID: Knowledge Base ID.
$SOURCE_ROW_ID: Source record ID.

Request example¶

curl -X POST "https://api.moi.matrixorigin.cn/v5/semantic-models/$MODEL_ID/sources/$SOURCE_ROW_ID/segments" \
  -H "X-API-Key: $AI_STUDIO_API_KEY" \
  -H "X-Workspace-ID: $WORKSPACE_ID" \
  -H 'Content-Type: application/json' \
  -d '{
    "if_match": "<DOCUMENT_ETAG>",
    "level": "chunk",
    "content": "<SEGMENT_CONTENT>"
  }'

Path parameters¶

Parameters	Type	Description
`model_id`	integer	Knowledge base ID.
`source_row_id`	string	Source record ID.

Request body¶

Submit the current snapshot’s ETag and the segment content. If the body has no nonempty if_match, supply the ETag through the If-Match header. At least one is required; a nonempty body value takes precedence.

Field	Type	Required	Description
`if_match`	string	Conditional	`data.etag` read before the call; may be omitted when using the `If-Match` header.
`level`	string	No	Segment level; defaults to `chunk` when omitted.
`content`	string	No	Text segment content.
`ocr_text`	string	No	OCR recognized text.
`image_description`	string	No	Image content description.
`bbox`	JSON	No	Boundary information for the segment in the original content.
`metadata`	object	No	Custom metadata object. Managed fields use the values written by the service.

The source must be a file source with a knowledge-base file ID, kb_file_id, and an available vector index; table sources do not support segment writes. Associated image and page-image IDs come from parsing. Submitting image_file_id or page_image_file_id during creation returns 400 with ErrParamInvalid.

Successful response¶

Returns 200 and the complete updated source document snapshot on success. The example below is a response excerpt.

{
  "code": "OK",
  "msg": "OK",
  "data": {
    "document": {
      "source": {
        "row_id": "src_01",
        "model_id": 401
      },
      "segment_status": {
        "available": true,
        "total": 3
      },
      "etag": "<NEW_DOCUMENT_ETAG>",
      "segments": [
        {
          "segment_id": "seg_03",
          "level": "chunk",
          "content": "新增分段",
          "enabled": true
        }
      ]
    }
  }
}

The response fields are as follows.

Field	Type	Description
`code`	string	`OK` on success.
`msg`	string	`OK` on success.
`data.document`	object	Snapshot of the source document after writing.
`data.document.source.row_id`	string	Source record ID.
`data.document.segment_status`	object	Current segment availability status and total number.
`data.document.etag`	string	ETag of the updated snapshot; returned when available. Use it for subsequent writes.
`data.document.segments[].segment_id`	string	New segment ID.
`data.document.segments[].level`	string	Segment level.
`data.document.segments[].content`	string	Segmented text; may be omitted if content is not returned.
`data.document.segments[].enabled`	boolean	Whether the segment participates in retrieval.

In field paths, [] means each item in an array. For example, data.document.segments[].segment_id identifies each segment in data.document.segments.

Error response¶

A missing or blank ETag returns 400 with ErrParamInvalid. An ETag that differs from the current snapshot returns 409 with ErrConflict. Error messages depend on the request language.

{
  "code": "ErrParamInvalid",
  "msg": "<ERROR_MESSAGE>",
  "data": null
}

Common HTTP errors¶

HTTP status code	error code	Common causes	Recommended actions
`400`	`ErrParamInvalid`	The path ID, request body, or ETag precondition is invalid.	Read the document again and supply its ETag using `if_match` or the `If-Match` header.
`401`	`ErrUnauthorized`	The API Key is invalid or has expired.	Check API Key.
`403`	`ErrForbidden`	The caller does not have permission to update this source.	Check workspace and object authorization.
`404`	`ErrNotFound`	The knowledge base or source does not exist or is not visible to the current caller.	Reread source list to confirm ID.
`409`	`ErrConflict`	The ETag does not match the current snapshot and the service refuses the write.	Read the latest snapshot and review the change before retrying with its ETag.
`500`	`ErrServer`	The service failed to create the segment.	Keep the desensitized response information and try again.

Follow-up operations¶

Read the document to confirm the new segment. Use the new snapshot’s ETag for subsequent writes; on 409 / ErrConflict, read the snapshot again and review the intended change.

Last updated on Oct 19, 2026