Parse (Local Files)

New

Upload local files directly and parse them into clean markdown. Use when the source document is a local file or not publicly accessible by URL. For documents available via a public URL, use the Scrape endpoint instead.

Overview

The /v2/parse endpoint accepts file uploads via multipart/form-data and processes them through Firecrawl's Rust-based parsing engine — up to 5x faster than URL-based document scraping. Supports PDF, DOCX, HTML, Excel, and more, with optional OCR, tag filtering, and zero data retention.

Endpoint

POST /v2/parse
Content-Type: multipart/form-data
Authorization: Bearer fc-YOUR-API-KEY

Supported File Types

Extension	Format	Max Size
.pdf	PDF document	50 MB
.docx, .doc	Microsoft Word	50 MB
.odt	OpenDocument Text	50 MB
.rtf	Rich Text Format	50 MB
.xlsx, .xls	Microsoft Excel	50 MB
.html, .htm	HTML document	50 MB

Request Fields

Required

Field	Type	Description
file	binary	The file to parse. Sent as a form-data field.

Optional options Object (JSON)

Pass as a JSON-encoded string in the options form field.

Option	Type	Default	Description
formats	array	[{type: "markdown"}]	Output types: markdown, summary, html, rawHtml, links, images, json
onlyMainContent	boolean	true	Exclude headers, navbars, and footers
includeTags	array	—	Only include these HTML tags in output
excludeTags	array	—	Strip these HTML tags from output
timeout	integer	30000	Timeout in milliseconds (max: 300000)
parsers	array	[{type: "pdf"}]	PDF parser config — see PDF Modes
removeBase64Images	boolean	true	Strip base64-encoded images from output
blockAds	boolean	true	Block ads and popups
skipTlsVerification	boolean	true	Skip TLS certificate verification
proxy	string	—	"basic" or "auto"
zeroDataRetention	boolean	false	Do not store file or output data after response

PDF Modes

Configure the parsers option to control how PDFs are processed:

Mode	Description	Use When
auto	Fast extraction first, falls back to OCR if needed	Default — works for most PDFs
fast	Text-only extraction, skips scanned pages	Digital-native PDFs with selectable text
ocr	Forces full OCR on every page	Scanned documents, image-heavy PDFs

"parsers": [{ "type": "pdf", "mode": "ocr", "maxPages": 20 }]

maxPages accepts 1–10000. Omit to process the entire document.

Examples

cURL

# Basic parse — returns markdown
curl -X POST 'https://api.firecrawl.dev/v2/parse' \
  -H 'Authorization: Bearer fc-YOUR-API-KEY' \
  -F 'file=@/path/to/document.pdf'

# With options — OCR mode, zero data retention
curl -X POST 'https://api.firecrawl.dev/v2/parse' \
  -H 'Authorization: Bearer fc-YOUR-API-KEY' \
  -F 'file=@/path/to/scanned-contract.pdf' \
  -F 'options={"parsers":[{"type":"pdf","mode":"ocr","maxPages":50}],"zeroDataRetention":true}'

# Request summary output as well
curl -X POST 'https://api.firecrawl.dev/v2/parse' \
  -H 'Authorization: Bearer fc-YOUR-API-KEY' \
  -F 'file=@/path/to/report.docx' \
  -F 'options={"formats":[{"type":"markdown"},{"type":"summary"}]}'

Python

import requests

api_key = "fc-YOUR-API-KEY"
url = "https://api.firecrawl.dev/v2/parse"

# Basic parse
with open("/path/to/document.pdf", "rb") as f:
    response = requests.post(
        url,
        headers={"Authorization": f"Bearer {api_key}"},
        files={"file": f}
    )

data = response.json()
print(data["data"]["markdown"])

import requests
import json

api_key = "fc-YOUR-API-KEY"
url = "https://api.firecrawl.dev/v2/parse"

options = {
    "parsers": [{"type": "pdf", "mode": "ocr", "maxPages": 100}],
    "formats": [{"type": "markdown"}, {"type": "summary"}],
    "zeroDataRetention": True
}

with open("/path/to/scanned-invoice.pdf", "rb") as f:
    response = requests.post(
        url,
        headers={"Authorization": f"Bearer {api_key}"},
        files={"file": f},
        data={"options": json.dumps(options)}
    )

result = response.json()
print(result["data"]["markdown"])
print(result["data"]["summary"])

Node.js

import fs from 'fs';
import FormData from 'form-data';
import fetch from 'node-fetch';

const apiKey = 'fc-YOUR-API-KEY';

// Basic parse
const form = new FormData();
form.append('file', fs.createReadStream('/path/to/document.pdf'));

const response = await fetch('https://api.firecrawl.dev/v2/parse', {
  method: 'POST',
  headers: {
    'Authorization': `Bearer ${apiKey}`,
    ...form.getHeaders()
  },
  body: form
});

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

import fs from 'fs';
import FormData from 'form-data';
import fetch from 'node-fetch';

const apiKey = 'fc-YOUR-API-KEY';

// With OCR options
const options = {
  parsers: [{ type: 'pdf', mode: 'ocr', maxPages: 50 }],
  zeroDataRetention: true
};

const form = new FormData();
form.append('file', fs.createReadStream('/path/to/scanned-doc.pdf'));
form.append('options', JSON.stringify(options));

const response = await fetch('https://api.firecrawl.dev/v2/parse', {
  method: 'POST',
  headers: {
    'Authorization': `Bearer ${apiKey}`,
    ...form.getHeaders()
  },
  body: form
});

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

Response

200 Success

{
  "success": true,
  "data": {
    "markdown": "string",
    "summary": "string (if requested)",
    "html": "string (if requested)",
    "rawHtml": "string (if requested)",
    "links": ["string array (if requested)"],
    "metadata": {
      "title": "string or array",
      "description": "string or array",
      "language": "string or array",
      "sourceURL": "uri",
      "url": "uri",
      "keywords": "string or array",
      "statusCode": 200,
      "contentType": "string",
      "error": null
    }
  }
}

Error Codes

Code	Meaning
400	Bad request — invalid file type, missing file field, or malformed options
402	Payment required — insufficient credits
429	Rate limited
500	Server error

Key Capabilities

• 5x faster — Rust-based parsing engine vs standard document scraping
• Up to 50 MB per file
• Zero data retention — set zeroDataRetention: true to prevent any storage of uploaded content or output
• OCR support — full optical character recognition for scanned documents
• Works offline — does not require the document to be publicly accessible

When to Use Parse vs Scrape

Scenario	Use
Document is a public URL	Scrape
Document is a local file	Parse (/v2/parse)
Document requires login to access	Parse (download first, then upload)
Need to parse many documents at once	Parse in a loop (one request per file)

---

Next: Document Parsing →