Get Upload URL

GET/api/v1/document/{extractionSchemaId}/get-url

Request a pre-signed URL for uploading a file to cloud storage. Returns both the upload URL and a fileId that you will use in subsequent processing requests.

Parameters

Name	Type	In	Required	Description
`extractionSchemaId`	string	path	Yes	The model's ID, the ID of a deployment that points at it, or the model's extraction schema ID. The extraction schema ID is returned when listing a project's deployments, and is also shown at the bottom of the data page in the dashboard.
`filename`	string	query	Yes	The original filename of the document. Used for file type detection and storage.
`allowDuplicate`	boolean	query	No	Set to true to skip duplicate detection and create a second document for content the project already stores. Omitting it is the same as false.

The platform hashes the uploaded content (SHA-256) and looks for a file in the same project with that hash whose status is neither FAILED nor UPLOADING. On a match no new document is created: the response carries the existing file's fileId and fileName, isDuplicate is true, and url is an empty string, so there is nothing to upload.

isDuplicate means the document is already stored in the project. It does not mean the document has been extracted. To extract a document that already exists, pass its fileId to one of the processing endpoints.

With allowDuplicate=true the hash lookup is skipped and a genuine second document is created, with a new fileId, a pre-signed upload URL, and isDuplicate set to false. The new document records the same content hash, so a later request with detection on matches the most recent copy.

Request

const response = await fetch(
`https://app.limai.io/api/v1/document/${SCHEMA_ID}/get-url?filename=${filename}`,
{
  headers: { "Authorization": `Bearer ${API_TOKEN}` }
}
)
const { url, fileId } = await response.json()

Response

Response200

{
  "url": "https://s3.amazonaws.com/bucket/path?AWSAccessKeyId=...",
  "fileId": "file_abc123def456",
  "fileName": "document.pdf",
  "isDuplicate": false,
  "modelId": "dep_abc123",
  "viaDeploymentId": "dpl_prod01"
}

Response Fields

Field	Type	Description
`url`	string	Pre-signed upload URL. Temporary and expires after a short period. Empty when the content is a duplicate.
`fileId`	string	Unique file identifier. Use this in subsequent API calls. For a duplicate, the identifier of the document the project already stores.
`fileName`	string	The stored document's name. Returned with a duplicate.
`isDuplicate`	boolean	Whether the content is already stored in the project.
`modelId`	string	The model the request resolved to.
`viaDeploymentId`	string \| null	The deployment the request arrived through, or `null` when the request named the model directly.

Duplicate Response

A duplicate response carries two extra fields naming where the stored document lives.

Response200

{
  "url": "",
  "fileId": "file_abc123def456",
  "fileName": "document.pdf",
  "isDuplicate": true,
  "deploymentId": "dep_abc123",
  "sameModel": true,
  "modelId": "dep_abc123",
  "viaDeploymentId": "dpl_prod01"
}

Field	Type	Description
`deploymentId`	string	The model the existing document belongs to. The lookup prefers the addressed model and otherwise falls back to any model in the project, so this is not always `modelId`.
`sameModel`	boolean	Whether the existing document belongs to the model this request addressed.

Error Responses

Filename missing:

Response400

{
  "error": "Missing filename query parameter"
}

allowDuplicate set to anything other than true or false: