Upload Documents
Upload multiple documents for asynchronous processing with vector embeddings.
This endpoint uploads documents and queues them for background processing using Celery. Each document is uploaded to S3 storage and processed to extract embeddings for semantic search and RAG applications. Only text-based file formats are supported.
Request Body (multipart/form-data):
files: List of files to upload (required)categories(optional): List of category names to organize documents
Returns:
list[int] containing:
- List of task IDs (integers) for tracking upload and processing status
- One task ID is returned for each uploaded file
- Use these IDs to monitor processing progress
Raises:
400: If no files provided400: If any file exceeds 100MB size limit400: If file type is unsupported (only text-based files allowed)400: If user doesn’t belong to an organization (required for vector processing)
Supported File Types:
Text-based files only: .txt, .md, .json, .jsonl, .xml, .csv, .html
File Size Limit:
Maximum 100 MB per file
Example Request:
POST /api/v3/documents
Headers: {"Authorization": "Bearer <api_key>"}
Content-Type: multipart/form-data
Form Data:
files: [report.txt, analysis.md]
categories: ["financial", "2024-q1"]
Example Response:
[
12345,
12346
]
Notes:
- Documents are processed asynchronously via Celery background tasks
- Processing includes S3 upload and vector embedding generation
- Organization membership is required for vector database access
- Files are validated for size and MIME type before processing
- Use task IDs to track processing status via appropriate status endpoints
Authorizations
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Body
Response
Returns a list of task IDs for the uploaded documents. This endpoint handles the upload of multiple documents, validates their types and sizes, creates processing tasks for each document, and returns their unique task identifiers for tracking the upload and processing status via Celery background tasks.