Layout Aware Parsing

The Document Parsing API parses a document, and returns a Markdown or JSON representation of the document.

You can use the markdown or JSON output to further perform any downstream processing.

• Chunking
• Summarization
• Image Indexing

​

Parsing a Document

curl -X 'POST' \
'https://api.tensorlake.ai/documents/v1/upload' \
-H 'accept: application/json' \
-H 'Content-Type: multipart/form-data' \
-H 'Authorization: Bearer YOUR_API_KEY' \
-F 'file_path=@/path/to/finance_stock_report.pdf;type=application/pdf'

curl -X GET https://api.tensorlake.ai/documents/v1/result \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
    "file": "tensorlake://b5dee680-c07c-4bad-ba00-7bd16d28975d"
}'

​

API Reference

​

Output Modes

The Document Parsing API supports two output modes:

• markdown
• json

Set the output mode using the output parameter of the /parse or /parse_async endpoint.

​

Switching between Models

You can switch between models using the model parameter of the /parse or /parse_async endpoint.

The following models are supported:

• small: Faster and smaller models are used for low latency parsing.
• medium: Slightly slower, uses combinations of vision and VLMs, but more accurate.
• large: Slowest, but most accurate, uses multiple VLMs and OCR models.

​

Listing Documents

You can list all the documents in a project using the following API call:

curl -X GET https://api.tensorlake.ai/documents/v1/list \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json"
-d '{
    "cursor": "xxx", 
    "limit": 10
}'