How to Get Started with the PDF.ai API

What this getting-started path is for

• You have an API key stored as a secret (shown only once when generated)

• You can call at least one endpoint (parse, then optionally extract, split, or ask) and get JSON back

• You know how credits work before you scale a batch

• Parse — PDF → markdown / structured contents (OCR via quality=standard)

• Extract — structured fields from a schema, with citations

• Split — divide a large PDF into smaller docs

• Ask — question a previously parsed docId (or several)

Practical rule: start on /developer for the key and plan. Use this guide for the first HTTP call. Keep the hub page for account and billing.

Browser product vs API

Credits to check before you scale

• Parse Standard (OCR): 1 credit/page; Advanced (VLM): 2 credits/page (extra may apply if image LLM is enabled)

• Extract: 2 credits/page (≤5 schema fields) or 4 credits/page (>5 fields), plus parse credits if the doc is not already cached

• Split: 2 credits/page (parse credits if not cached)

• Ask: 3 credits × total page count across the docIds you query (documents must be parsed first)

• Delete: no cost

• Cached parse with the same settings: 0 credits

• Free: $0/month for 200 credits/month, no credit card required

• Paid examples: $49 / 3,000, $99 / 10,000, $249 / 30,000, $599 / 100,000 credits per month

Practical rule: one successful parse on a short PDF proves the key. Do not burn a large batch until credit math matches your page counts.

How to make the first request

1. Create an API key

2. Parse a PDF (recommended first call)

• file or url (or docId when you already have a cached document id)

• quality: standard (default, OCR path) or advanced (VLM)

• lang_list: languages for standard OCR (default ["en"])

• llm: optional image LLM processing (default false)

import requests

url = "https://pdf.ai/api/v2/parse"
headers = {"X-API-Key": "YOUR_API_KEY"}

with open("/path/to/document.pdf", "rb") as f:
    files = {"file": f}
    data = {
        "quality": "standard",
        "lang_list": '["en"]'
    }
    response = requests.post(url, headers=headers, files=files, data=data)

print(response.json())

3. Optional next calls

4. Check the output before you trust it

• Confirm success and that pageCount matches the file you sent

• For parse: search the markdown for a word from page 1 and a later page you care about

• For ask: open the section the answer claims to use before you ship it downstream

• If you get Invalid API key / No API key present, fix the header before you change the PDF

Practical rule: a non-empty JSON body is not a verified document. Read the lines you will use.

When this path is the wrong tool

• You only need one chat answer in a browser. Use the web app path in How to Chat with a PDF. No API key required for that job.

• You only need one searchable PDF. Use https://pdf.ai/tools/ocr-pdf, then prove search works.

• You wanted a generic “pdf api” category hub. Stay on https://pdf.ai/developer. This page is the first-request how-to, not a replacement hub.

• You need OCR-in-code detail only. Prefer PDF OCR API: Run OCR from Your Code (quality=standard on Parse).

A short action plan

1. Open https://pdf.ai/developer, create a key, store it once.

2. Check credit usage against your page count.

3. POST https://pdf.ai/api/v2/parse with quality=standard on a short PDF.

4. Save docId. Optionally call Ask or Extract next.

5. Verify text or answers against the source PDF before you wire production.