API DOCUMENTATION

Nural+ Extractor API

Convert any document into clean, structured JSON. Upload PDFs, Word docs, spreadsheets, presentations, images — and get normalized, chunked content ready for LLMs and AI pipelines.

Base URL

https://neural-extarctor.netlify.app

Version

v1

Auth

None required

Endpoint

POST

/api/v1/extract

Upload one or more files and receive structured extraction results as JSON. Files are processed in parallel with per-file error isolation.

📦

Max File Size

50 MB per file

⏱️

Max Duration

60 seconds

📁

Batch Upload

Multiple files per request

🔓

CORS

All origins allowed

Request

Headers

Header	Value	Required
`Content-Type`	`multipart/form-data`	Required

Form Fields

Field	Type	Description
`files`	File(s)	One or more files to extract. Use the same field name for batch uploads.

Response

Top-Level Fields

Field	Type	Description
`apiVersion`	`string`	Always "v1"
`timestamp`	`string`	ISO-8601 timestamp of the response
`totalFiles`	`number`	Number of files submitted
`successful`	`number`	Number of successfully extracted files
`failed`	`number`	Number of failed extractions
`results`	`ExtractionResult[]`	Per-file extraction results

Example Response

json

{
  "apiVersion": "v1",
  "timestamp": "2026-03-02T12:00:00.000Z",
  "totalFiles": 2,
  "successful": 1,
  "failed": 1,
  "results": [
    {
      "success": true,
      "document": {
        "documentId": "550e8400-e29b-41d4-a716-446655440000",
        "fileName": "report.pdf",
        "mimeType": "application/pdf",
        "metadata": { "pageCount": 12 },
        "chunks": [
          {
            "id": "6ba7b810-9dad-11d1-80b4-00c04fd430c8",
            "type": "heading",
            "text": "Introduction",
            "page": 1
          },
          {
            "id": "6ba7b811-9dad-11d1-80b4-00c04fd430c8",
            "type": "paragraph",
            "text": "This report covers the quarterly results...",
            "page": 1,
            "section": "Introduction"
          }
        ]
      }
    },
    {
      "success": false,
      "fileName": "corrupt.pdf",
      "error": "Failed to parse PDF: Invalid header"
    }
  ]
}

Schema

ExtractedDocument

Field	Type	Description
`documentId`	`string (UUID)`	Unique identifier for this extraction
`fileName`	`string`	Original file name
`mimeType`	`string`	Detected MIME type
`metadata`	`object`	File-level metadata (pageCount, sheetCount, etc.)
`chunks`	`ContentChunk[]`	Ordered list of content chunks

ContentChunk

Field	Type	Description
`id`	`string (UUID)`	Unique chunk identifier
`type`	`string`	One of: "heading", "paragraph", "table", "list"
`text`	`string`	Extracted and normalized text content
`page`	`number?`	Source page number (1-indexed), if applicable
`section`	`string?`	Logical section or slide name, if available

Supported Formats

Format	Extensions	MIME Type	Library
PDF	`.pdf`	`application/pdf`	pdfjs-serverless
Word	`.docx, .doc`	`application/vnd.openxml...document`	mammoth
Excel	`.xlsx, .xls`	`application/vnd.openxml...sheet`	exceljs
CSV / TSV	`.csv, .tsv`	`text/csv, text/tab-separated-values`	csv-parse
PowerPoint	`.pptx, .ppt`	`application/vnd.openxml...presentation`	jszip
HTML	`.html, .htm`	`text/html`	built-in
Images	`.png, .jpg, .jpeg, .webp, .tiff, .gif, .bmp`	`image/*`	sharp + Tesseract.js

Examples

cURL — Single File

bash

curl -X POST https://neural-extarctor.netlify.app/api/v1/extract \
  -F "files=@report.pdf"

cURL — Batch Upload

bash

curl -X POST https://neural-extarctor.netlify.app/api/v1/extract \
  -F "files=@report.pdf" \
  -F "files=@data.xlsx" \
  -F "files=@notes.docx"

JavaScript / Fetch

javascript

const formData = new FormData();
formData.append("files", fileInput.files[0]);

const response = await fetch("/api/v1/extract", {
  method: "POST",
  body: formData,
});

const data = await response.json();
console.log(data.results);

Python

python

import requests

url = "https://neural-extarctor.netlify.app/api/v1/extract"
files = [
    ("files", ("report.pdf", open("report.pdf", "rb"), "application/pdf")),
    ("files", ("data.xlsx", open("data.xlsx", "rb"),
     "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet")),
]

response = requests.post(url, files=files)
print(response.json())

GET — Service Discovery

bash

curl https://neural-extarctor.netlify.app/api/v1/extract

The GET endpoint returns supported formats, limits, and usage info — useful for health checks and integration discovery.

Error Handling

Status	Condition	Description
200	Success (even partial)	All or some files extracted. Check results[i].success per file.
400	No files provided	Request had no files in the 'files' form field.
415	Wrong content type	Expected multipart/form-data.
500	Server error	Unhandled internal error.

💡

Per-file isolation: Even on a 200 response, individual files may have failed. Always check results[i].success for each file in the response.

🔍

MIME fallback: If the browser sends application/octet-stream, the API automatically detects MIME type from the file extension.