How to Understand PDF.ai API Credit Usage

Publish date

Sep 25, 2026

AI summary

A concise guide explains PDF.ai API credit usage, detailing how Parse, Extract, Split, Ask, and Delete endpoints consume credits based on page count, quality, image processing, and schema size, and shows how cached docIds can eliminate repeat Parse costs; it also outlines free and paid plan credit limits and provides practical steps to estimate workflow credits and compare them to plan allowances.

Language

People look up API credits when they already know the endpoint they need, but not how many credits a 10-page file will burn across Parse, Extract, Split, or Ask. They want the math before they scale page count.

This page is the credit + plan companion for the PDF.ai API v2 docs. It does not replace the API Hub (keys and Manage Plan live there). It does not re-rank the head term pdf api. For the first-request path (key + first call), see How to Get Started with the PDF.ai API. Endpoint how-tos: Parse, Extract, Split, Ask.

What you should know after this page

After you read the live credit usage docs and the per-endpoint pages, three things should be true:

You can estimate credits for Parse, Extract, Split, Ask, and Delete from page count (and schema size for Extract)

You know when a cached docId makes Parse credits 0 on Extract or Split

You know the Free plan is $0/month for 200 credits/month, with no credit card required, and where paid plan examples live

Docs: Credit usage.

Practical rule: Parse once, keep the docId, then Extract / Split / Ask against that cache so you are not paying parse credits on every follow-up call.

Credit rates by endpoint

Numbers below come from the live endpoint docs and the credit usage summary. Prefer the endpoint page when a detail conflicts with a short summary line.

Parse (`POST /v2/parse`)

From Parse:

Quality	Base cost per page	LLM cost (if enabled)
`standard` (default)	1 credit	+1 credit per 5 images (rounded up)
`advanced`	2 credits	+1 credit per 5 images (rounded up)

Cached results (same docId + same settings) use 0 credits.

Examples from the Parse docs:

10-page document, standard, no LLM → 10 credits

10-page document, advanced, no LLM → 20 credits

10-page document, standard, LLM on, 12 images → base 10 + LLM 3 = 13 credits

Cached result → 0 credits

Extract (`POST /v2/extract`)

From Extract: the document must be parsed first. Parse credits apply unless you pass a cached docId.

Component	Condition	Calculation
Extraction credits	Schema has ≤ 5 fields	2 credits × page count
Extraction credits	Schema has > 5 fields	4 credits × page count

Total = Parse credits + Extraction credits

Examples from the Extract docs:

10-page PDF, not cached, advanced parse, schema with 3 fields → parse 20 + extract 20 = 40 credits

5-page PDF, cached parse, schema with 8 fields → parse 0 + extract 20 = 20 credits

Note: the short credit usage page lists Extract as “Standard 2 / Advanced 4” per page. The Extract endpoint page defines extraction cost by schema field count (≤5 vs >5). Use the endpoint formula when you are sizing an Extract call.

Split (`POST /v2/split`)

From Split:

Split credits: always 2 credits × page count

Parse credits: charged when you send a file or url; 0 when you pass a cached docId

Total = Parse credits + Split credits

Examples from the Split docs:

10-page PDF, cached docId → parse 0 + split 20 = 20 credits

5-page PDF, not cached, advanced quality → parse 10 + split 10 = 20 credits

Ask (`POST /v2/ask`)

From Ask: documents in docIds must already be parsed.

Ask credits: always 3 credits × total page count across all documents in the request

Total credits = total page count × 3

Example from the Ask docs: Document 1 = 10 pages, Document 2 = 5 pages → (10 + 5) × 3 = 45 credits.

Delete (`POST /v2/delete`)

From Delete and credit usage:

Free — no API credits for delete

Permanent for v2 documents created via parse, extract, or split

Quick reference

Endpoint	Typical credit rule	Needs prior Parse?
Parse	1 or 2 per page by quality (+ optional LLM image blocks)	No
Extract	2 or 4 per page by schema size + Parse if uncached	Yes (or parse during the call)
Split	2 per page + Parse if uncached	Yes (or parse during the call)
Ask	3 × total pages across `docIds`	Yes (`docIds` must be parsed)
Delete	0	N/A

API plan examples (from the docs)

From credit usage. Generate the key and manage the plan on https://pdf.ai/developer (Manage Plan). The Quick start docs say the key is shown only once when generated.

Free plan

$0/month for 200 credits per month

No credit card required to get started

Paid plan examples

$49/month for 3,000 credits/month ($0.016/credit)

$99/month for 10,000 credits/month ($0.010/credit)

$249/month for 30,000 credits/month ($0.008/credit)

$599/month for 100,000 credits/month ($0.006/credit)

Contact support@pdf.ai for more than 100,000 credits per month

Practical rule: size a short PDF on Free (200 credits/month) before you commit page volume on a paid tier.

How to estimate a workflow

1. Count pages and pick quality

Decide standard vs advanced for any call that may run Parse. Advanced is 2 credits/page for Parse; standard is 1.

2. Prefer a cached `docId`

Run Parse once (or accept parse cost on the first Extract/Split with file/url). Save docId. Later Extract / Split / Ask calls that reuse it avoid repeat parse charges (Ask still needs a prior parse; Extract/Split skip parse credits when docId is passed).

3. Add the endpoint layer

Extract: 2× or 4× pages from schema field count

Split: 2× pages

Ask: 3× total pages across all docIds

4. Compare to plan credits

Check Free 200/month or a paid example on credit usage against your estimate before you batch large files.

When this page is the wrong tool

You need a key and a first Parse call. Use How to Get Started with the PDF.ai API.

You need endpoint request shapes. Use the Parse / Extract / Split / Ask how-tos linked above, or the live docs on api.pdf.ai.

You wanted a generic “pdf api” category hub. Stay on https://pdf.ai/developer. This page is the credit companion, not a replacement hub.

You only need a browser tool, not API credits. Use the free tools such as Split PDF when the job does not need code.

A short action plan

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

Read credit usage and the endpoint page for the call you will make.

Parse a short PDF, save docId, then run Extract, Split, or Ask against that cache.

Add up credits from the tables above and compare to Free 200/month or a paid example before you scale.

Sibling guides: How to Get Started with the PDF.ai API, How to Use the PDF.ai Parse API, How to Use the PDF.ai Extract API, How to Use the PDF.ai Split API, How to Use the PDF.ai Ask API, PDF OCR API: Run OCR from Your Code.

If you want the rest of the PDF workflow after credits are clear, start at PDF.ai. Use these rates when the next step is your pipeline budget, not a single upload box.