Imported from khanhduyvt0101/pdfvector-skills (skills/pdfvector-api/SKILL.md). Install upstream with npx skills add khanhduyvt0101/pdfvector-skills --skill pdfvector-api. Copyright stays with the author.

PDF Vector API

Use this skill when a user wants to call PDF Vector public APIs from an agent or app. It covers document, invoice, identity, bank statement, and academic APIs.

Sources

API reference: https://global.pdfvector.com/api/reference
TypeScript SDK: https://www.npmjs.com/package/@pdfvector/client
Website: https://www.pdfvector.com

Authentication

Use a PDF Vector API key in the Authorization header:

Authorization: Bearer YOUR_API_KEY

Never print, log, commit, or echo a real API key. Use YOUR_API_KEY in examples.

TypeScript SDK

Install the official client when the runtime is JavaScript or TypeScript:

npm install @pdfvector/client

Create a client:

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

For custom or self-hosted instances, pass the instance domain:

const client = createClient({
  domain: "global.pdfvector.com",
  apiKey: "YOUR_API_KEY",
});

Page-Separated Output

Set includePages: true when the caller needs page-separated markdown. PDF Vector still returns the full document in top-level markdown.

Relevant response fields:

markdown: full parsed markdown for the document or paper.
pages: page-level markdown array, available when includePages is true.
pages[].pageNumber: 1-based page number.
pages[].markdown: markdown for that page.
pageCount: total pages.
model: model tier used.
credits: credits charged.
requestId: request identifier for support/debugging.
html: full HTML for Document Parse when using the max model.

Credit Costs

API	nano	mini	pro	max	Unit
Document Parse	1	2	4	8	/page
Document Ask	2	4	8	16	/page
Document Extract	2	4	8	16	/page
Identity Parse	-	-	6	10	/page
Identity Ask	6	10	14	18	/page
Identity Extract	6	10	14	18	/page
Invoice Parse	-	-	6	10	/page
Invoice Ask	6	10	14	18	/page
Invoice Extract	6	10	14	18	/page
Bank Statement Parse	-	-	6	10	/page
Bank Statement Ask	6	10	14	18	/page
Bank Statement Extract	6	10	14	18	/page
Academic Search	2	2	2	2	/request
Academic Fetch	2	2	2	2	/request
Academic Parse	1	2	4	8	/page
Academic Find Citations	2	2	2	2	/sentence
Academic Paper Graph	2+	2+	2+	2+	/request
Academic Similar Papers	3	3	3	3	/request
Academic Search Grants	2	2	2	2	/request

Model Guidance

Use model: "auto" by default. Other model tiers are "nano", "mini", "pro", and "max".

Choose higher tiers when the document is complex, layout-heavy, or needs stronger extraction quality. Use max for Document Parse when the caller needs HTML output.

Agent Workflow

Ask for or locate the PDF Vector API key without exposing it.
Choose document.parse for direct documents or academic.parse for papers and publication IDs.
Set includePages: true when the user asks for page-by-page output, citations by page, page summaries, or chunking by page.
Return markdown for the full result and pages when the caller needs page-level content.
Include requestId in debugging notes when a request fails.

Public API Examples

Document Parse

cURL

curl -X POST https://global.pdfvector.com/api/document/parse \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: my-doc-123" \
  -d '{
    "url": "https://example.com/annual-report.pdf",
    "model": "max",
    "includePages": true
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.document.parse(
  { url: "https://example.com/annual-report.pdf", model: "max", includePages: true },
  { context: { documentId: "my-doc-123" } },
);

console.log(result.markdown);
console.log(result.pages?.[0]?.markdown);
console.log(`Pages: ${result.pageCount}, Model: ${result.model}`);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/document/parse",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "my-doc-123",
    },
    json={
        "url": "https://example.com/annual-report.pdf",
        "model": "max",
        "includePages": True,
    },
)
data = response.json()
print(data["markdown"], data["pages"][0]["markdown"], data["pageCount"])

Example Response

{
  "documentId": "my-doc-123",
  "markdown": "# Annual Report 2024\n\n## Executive Summary\nRevenue grew 23% year-over-year to $4.2B...",
  "pageCount": 5,
  "pages": [{ "pageNumber": 1, "markdown": "# Annual Report 2024\n\n## Executive Summary..." }],
  "model": "max",
  "html": "<h1>Annual Report 2024</h1><h2>Executive Summary</h2><p>Revenue grew 23% year-over-year to $4.2B...</p>",
  "credits": 40,
  "requestId": 42
}

Document Ask

cURL

curl -X POST https://global.pdfvector.com/api/document/ask \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: my-doc-123" \
  -d '{
    "url": "https://example.com/annual-report.pdf",
    "question": "What was the total revenue in Q4?",
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.document.ask(
  {
    url: "https://example.com/annual-report.pdf",
    question: "What was the total revenue in Q4?",
    model: "max",
  },
  { context: { documentId: "my-doc-123" } },
);

console.log(result.markdown);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/document/ask",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "my-doc-123",
    },
    json={
        "url": "https://example.com/annual-report.pdf",
        "question": "What was the total revenue in Q4?",
        "model": "max",
    },
)
data = response.json()
print(data["markdown"])

Example Response

{
  "documentId": "my-doc-123",
  "markdown": "Based on the document, the total revenue in Q4 was $1.12B, representing a 15% increase from Q3 and a 23% year-over-year growth.",
  "model": "max",
  "requestId": 43
}

Document Extract

cURL

curl -X POST https://global.pdfvector.com/api/document/extract \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: my-doc-123" \
  -d '{
    "url": "https://example.com/annual-report.pdf",
    "prompt": "Extract the company overview, fiscal year, revenue, net income, and key business segments",
    "schema": {
      "type": "object",
      "properties": {
        "companyName": { "type": "string" },
        "fiscalYear": { "type": "string" },
        "totalRevenue": { "type": "number" },
        "netIncome": { "type": "number" },
        "segments": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "name": { "type": "string" },
              "revenue": { "type": "number" },
              "growthPercent": { "type": "number" }
            }
          }
        }
      }
    },
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.document.extract(
  {
    url: "https://example.com/annual-report.pdf",
    prompt: "Extract the company overview, fiscal year, revenue, net income, and key business segments",
    schema: {
      type: "object",
      properties: {
        companyName: { type: "string" },
        fiscalYear: { type: "string" },
        totalRevenue: { type: "number" },
        netIncome: { type: "number" },
        segments: {
          type: "array",
          items: {
            type: "object",
            properties: {
              name: { type: "string" },
              revenue: { type: "number" },
              growthPercent: { type: "number" },
            },
          },
        },
      },
    },
    model: "max",
  },
  { context: { documentId: "my-doc-123" } },
);

console.log(result.data);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/document/extract",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "my-doc-123",
    },
    json={
        "url": "https://example.com/annual-report.pdf",
        "prompt": "Extract the company overview, fiscal year, revenue, net income, and key business segments",
        "schema": {
            "type": "object",
            "properties": {
                "companyName": {"type": "string"},
                "fiscalYear": {"type": "string"},
                "totalRevenue": {"type": "number"},
                "netIncome": {"type": "number"},
                "segments": {
                    "type": "array",
                    "items": {
                        "type": "object",
                        "properties": {
                            "name": {"type": "string"},
                            "revenue": {"type": "number"},
                            "growthPercent": {"type": "number"},
                        },
                    },
                },
            },
        },
        "model": "max",
    },
)
data = response.json()
print(data["data"])

Example Response

{
  "documentId": "my-doc-123",
  "data": {
    "companyName": "Acme Corporation",
    "fiscalYear": "2024",
    "totalRevenue": 4200000000,
    "netIncome": 820000000,
    "segments": [
      { "name": "Cloud Services", "revenue": 2100000000, "growthPercent": 34 },
      { "name": "Enterprise Software", "revenue": 1400000000, "growthPercent": 18 },
      { "name": "Consulting", "revenue": 700000000, "growthPercent": 12 }
    ]
  },
  "model": "max",
  "requestId": 44
}

Identity Parse

cURL

curl -X POST https://global.pdfvector.com/api/identity/parse \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: passport-456" \
  -d '{
    "url": "https://example.com/passport-scan.jpg",
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.identity.parse(
  { url: "https://example.com/passport-scan.jpg", model: "max" },
  { context: { documentId: "passport-456" } },
);

console.log(result.markdown);
console.log(`Type: ${result.documentType}, Model: ${result.model}`);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/identity/parse",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "passport-456",
    },
    json={
        "url": "https://example.com/passport-scan.jpg",
        "model": "max",
    },
)
data = response.json()
print(data["markdown"])

Example Response

{
  "documentId": "passport-456",
  "markdown": "# Passport — John A. Smith\n\n**Full Name:** John Alexander Smith\n**Nationality:** United States\n**Date of Birth:** 1985-03-15\n**Expiry Date:** 2029-07-22",
  "pageCount": 1,
  "model": "max",
  "html": "<h1>Passport — John A. Smith</h1><p><strong>Full Name:</strong> John Alexander Smith</p><p><strong>Nationality:</strong> United States</p>",
  "documentType": "passport",
  "requestId": 42
}

Identity Ask

cURL

curl -X POST https://global.pdfvector.com/api/identity/ask \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: passport-456" \
  -d '{
    "url": "https://example.com/passport-scan.jpg",
    "question": "What is the expiry date?",
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.identity.ask(
  {
    url: "https://example.com/passport-scan.jpg",
    question: "What is the expiry date?",
    model: "max",
  },
  { context: { documentId: "passport-456" } },
);

console.log(result.markdown);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/identity/ask",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "passport-456",
    },
    json={
        "url": "https://example.com/passport-scan.jpg",
        "question": "What is the expiry date?",
        "model": "max",
    },
)
data = response.json()
print(data["markdown"])

Example Response

{
  "documentId": "passport-456",
  "markdown": "The expiry date on the passport is July 22, 2029.",
  "model": "max",
  "requestId": 43
}

Identity Extract

cURL

curl -X POST https://global.pdfvector.com/api/identity/extract \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: passport-456" \
  -d '{
    "url": "https://example.com/passport-scan.jpg",
    "prompt": "Extract the full name, date of birth, nationality, passport number, and expiry date",
    "schema": {
      "type": "object",
      "properties": {
        "fullName": { "type": "string" },
        "dateOfBirth": { "type": "string" },
        "nationality": { "type": "string" },
        "passportNumber": { "type": "string" },
        "expiryDate": { "type": "string" }
      }
    },
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.identity.extract(
  {
    url: "https://example.com/passport-scan.jpg",
    prompt: "Extract the full name, date of birth, nationality, passport number, and expiry date",
    schema: {
      type: "object",
      properties: {
        fullName: { type: "string" },
        dateOfBirth: { type: "string" },
        nationality: { type: "string" },
        passportNumber: { type: "string" },
        expiryDate: { type: "string" },
      },
    },
    model: "max",
  },
  { context: { documentId: "passport-456" } },
);

console.log(result.data);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/identity/extract",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "passport-456",
    },
    json={
        "url": "https://example.com/passport-scan.jpg",
        "prompt": "Extract the full name, date of birth, nationality, passport number, and expiry date",
        "schema": {
            "type": "object",
            "properties": {
                "fullName": {"type": "string"},
                "dateOfBirth": {"type": "string"},
                "nationality": {"type": "string"},
                "passportNumber": {"type": "string"},
                "expiryDate": {"type": "string"},
            },
        },
        "model": "max",
    },
)
data = response.json()
print(data["data"])

Example Response

{
  "documentId": "passport-456",
  "data": {
    "fullName": "John Alexander Smith",
    "dateOfBirth": "1985-03-15",
    "nationality": "United States",
    "passportNumber": "E12345678",
    "expiryDate": "2029-07-22"
  },
  "model": "max",
  "requestId": 44
}

Invoice Parse

cURL

curl -X POST https://global.pdfvector.com/api/invoice/parse \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: inv-2024-0892" \
  -d '{
    "url": "https://example.com/supplier-invoice.pdf",
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.invoice.parse(
  { url: "https://example.com/supplier-invoice.pdf", model: "max" },
  { context: { documentId: "inv-2024-0892" } },
);

console.log(result.markdown);
console.log(`Pages: ${result.pageCount}, Model: ${result.model}`);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/invoice/parse",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "inv-2024-0892",
    },
    json={
        "url": "https://example.com/supplier-invoice.pdf",
        "model": "max",
    },
)
data = response.json()
print(data["markdown"])

Example Response

{
  "documentId": "inv-2024-0892",
  "markdown": "# Invoice #INV-2024-0892\n\n**Vendor:** TechSupply Co.\n**Date:** 2024-11-15\n**Due Date:** 2024-12-15\n\n| Item | Qty | Price |\n|------|-----|-------|\n| Server Rack | 2 | $3,400.00 |\n| Network Switch | 4 | $1,200.00 |\n\n**Total: $4,600.00**",
  "pageCount": 1,
  "model": "max",
  "html": "<h1>Invoice #INV-2024-0892</h1><p><strong>Vendor:</strong> TechSupply Co.</p><table><tr><td>Server Rack</td><td>$3,400.00</td></tr></table>",
  "requestId": 42
}

Invoice Ask

cURL

curl -X POST https://global.pdfvector.com/api/invoice/ask \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: inv-2024-0892" \
  -d '{
    "url": "https://example.com/supplier-invoice.pdf",
    "question": "What is the total amount due?",
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.invoice.ask(
  {
    url: "https://example.com/supplier-invoice.pdf",
    question: "What is the total amount due?",
    model: "max",
  },
  { context: { documentId: "inv-2024-0892" } },
);

console.log(result.markdown);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/invoice/ask",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "inv-2024-0892",
    },
    json={
        "url": "https://example.com/supplier-invoice.pdf",
        "question": "What is the total amount due?",
        "model": "max",
    },
)
data = response.json()
print(data["markdown"])

Example Response

{
  "documentId": "inv-2024-0892",
  "markdown": "The total amount due on invoice #INV-2024-0892 is $4,600.00, with a payment due date of December 15, 2024.",
  "model": "max",
  "requestId": 43
}

Invoice Extract

cURL

curl -X POST https://global.pdfvector.com/api/invoice/extract \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: inv-2024-0892" \
  -d '{
    "url": "https://example.com/supplier-invoice.pdf",
    "prompt": "Extract vendor name, invoice number, date, total amount, and all line items",
    "schema": {
      "type": "object",
      "properties": {
        "vendor": { "type": "string" },
        "invoiceNumber": { "type": "string" },
        "date": { "type": "string" },
        "total": { "type": "number" },
        "lineItems": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "description": { "type": "string" },
              "quantity": { "type": "number" },
              "unitPrice": { "type": "number" },
              "amount": { "type": "number" }
            }
          }
        }
      }
    },
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.invoice.extract(
  {
    url: "https://example.com/supplier-invoice.pdf",
    prompt: "Extract vendor name, invoice number, date, total amount, and all line items",
    schema: {
      type: "object",
      properties: {
        vendor: { type: "string" },
        invoiceNumber: { type: "string" },
        date: { type: "string" },
        total: { type: "number" },
        lineItems: {
          type: "array",
          items: {
            type: "object",
            properties: {
              description: { type: "string" },
              quantity: { type: "number" },
              unitPrice: { type: "number" },
              amount: { type: "number" },
            },
          },
        },
      },
    },
    model: "max",
  },
  { context: { documentId: "inv-2024-0892" } },
);

console.log(result.data);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/invoice/extract",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "inv-2024-0892",
    },
    json={
        "url": "https://example.com/supplier-invoice.pdf",
        "prompt": "Extract vendor name, invoice number, date, total amount, and all line items",
        "schema": {
            "type": "object",
            "properties": {
                "vendor": {"type": "string"},
                "invoiceNumber": {"type": "string"},
                "date": {"type": "string"},
                "total": {"type": "number"},
                "lineItems": {
                    "type": "array",
                    "items": {
                        "type": "object",
                        "properties": {
                            "description": {"type": "string"},
                            "quantity": {"type": "number"},
                            "unitPrice": {"type": "number"},
                            "amount": {"type": "number"},
                        },
                    },
                },
            },
        },
        "model": "max",
    },
)
data = response.json()
print(data["data"])

Example Response

{
  "documentId": "inv-2024-0892",
  "data": {
    "vendor": "TechSupply Co.",
    "invoiceNumber": "INV-2024-0892",
    "date": "2024-11-15",
    "total": 4600.00,
    "lineItems": [
      { "description": "Server Rack", "quantity": 2, "unitPrice": 1700.00, "amount": 3400.00 },
      { "description": "Network Switch", "quantity": 4, "unitPrice": 300.00, "amount": 1200.00 }
    ]
  },
  "model": "max",
  "requestId": 44
}

Bank Statement Parse

cURL

curl -X POST https://global.pdfvector.com/api/bankStatement/parse \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: stmt-nov-2024" \
  -d '{
    "url": "https://example.com/monthly-statement.pdf",
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.bankStatement.parse(
  { url: "https://example.com/monthly-statement.pdf", model: "max" },
  { context: { documentId: "stmt-nov-2024" } },
);

console.log(result.markdown);
console.log(`Pages: ${result.pageCount}, Model: ${result.model}`);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/bankStatement/parse",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "stmt-nov-2024",
    },
    json={
        "url": "https://example.com/monthly-statement.pdf",
        "model": "max",
    },
)
data = response.json()
print(data["markdown"])

Example Response

{
  "documentId": "stmt-nov-2024",
  "markdown": "# Bank Statement — Chase Checking ****4821\n\n**Period:** Nov 1 – Nov 30, 2024\n**Opening Balance:** $12,450.33\n**Closing Balance:** $14,892.17\n\n| Date | Description | Amount |\n|------|-------------|--------|\n| 11/02 | Direct Deposit | +$5,200.00 |\n| 11/05 | Rent Payment | -$2,100.00 |",
  "pageCount": 3,
  "model": "max",
  "html": "<h1>Bank Statement — Chase Checking ****4821</h1><table><tr><td>11/02</td><td>Direct Deposit</td><td>+$5,200.00</td></tr></table>",
  "requestId": 42
}

Bank Statement Ask

cURL

curl -X POST https://global.pdfvector.com/api/bankStatement/ask \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: stmt-nov-2024" \
  -d '{
    "url": "https://example.com/monthly-statement.pdf",
    "question": "What is the closing balance?",
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.bankStatement.ask(
  {
    url: "https://example.com/monthly-statement.pdf",
    question: "What is the closing balance?",
    model: "max",
  },
  { context: { documentId: "stmt-nov-2024" } },
);

console.log(result.markdown);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/bankStatement/ask",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "stmt-nov-2024",
    },
    json={
        "url": "https://example.com/monthly-statement.pdf",
        "question": "What is the closing balance?",
        "model": "max",
    },
)
data = response.json()
print(data["markdown"])

Example Response

{
  "documentId": "stmt-nov-2024",
  "markdown": "The closing balance for the statement period ending November 30, 2024 is $14,892.17.",
  "model": "max",
  "requestId": 43
}

Bank Statement Extract

cURL

curl -X POST https://global.pdfvector.com/api/bankStatement/extract \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-pdfvector-document-id: stmt-nov-2024" \
  -d '{
    "url": "https://example.com/monthly-statement.pdf",
    "prompt": "Extract account holder, opening balance, closing balance, statement period, and all transactions",
    "schema": {
      "type": "object",
      "properties": {
        "accountHolder": { "type": "string" },
        "openingBalance": { "type": "number" },
        "closingBalance": { "type": "number" },
        "statementPeriod": { "type": "string" },
        "transactions": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "date": { "type": "string" },
              "description": { "type": "string" },
              "amount": { "type": "number" },
              "balance": { "type": "number" }
            }
          }
        }
      }
    },
    "model": "max"
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.bankStatement.extract(
  {
    url: "https://example.com/monthly-statement.pdf",
    prompt: "Extract account holder, opening balance, closing balance, statement period, and all transactions",
    schema: {
      type: "object",
      properties: {
        accountHolder: { type: "string" },
        openingBalance: { type: "number" },
        closingBalance: { type: "number" },
        statementPeriod: { type: "string" },
        transactions: {
          type: "array",
          items: {
            type: "object",
            properties: {
              date: { type: "string" },
              description: { type: "string" },
              amount: { type: "number" },
              balance: { type: "number" },
            },
          },
        },
      },
    },
    model: "max",
  },
  { context: { documentId: "stmt-nov-2024" } },
);

console.log(result.data);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/bankStatement/extract",
    headers={
        "Authorization": "Bearer YOUR_API_KEY",
        "x-pdfvector-document-id": "stmt-nov-2024",
    },
    json={
        "url": "https://example.com/monthly-statement.pdf",
        "prompt": "Extract account holder, opening balance, closing balance, statement period, and all transactions",
        "schema": {
            "type": "object",
            "properties": {
                "accountHolder": {"type": "string"},
                "openingBalance": {"type": "number"},
                "closingBalance": {"type": "number"},
                "statementPeriod": {"type": "string"},
                "transactions": {
                    "type": "array",
                    "items": {
                        "type": "object",
                        "properties": {
                            "date": {"type": "string"},
                            "description": {"type": "string"},
                            "amount": {"type": "number"},
                            "balance": {"type": "number"},
                        },
                    },
                },
            },
        },
        "model": "max",
    },
)
data = response.json()
print(data["data"])

Example Response

{
  "documentId": "stmt-nov-2024",
  "data": {
    "accountHolder": "John A. Smith",
    "openingBalance": 12450.33,
    "closingBalance": 14892.17,
    "statementPeriod": "Nov 1 – Nov 30, 2024",
    "transactions": [
      { "date": "2024-11-02", "description": "Direct Deposit - Employer", "amount": 5200.00, "balance": 17650.33 },
      { "date": "2024-11-05", "description": "Rent Payment", "amount": -2100.00, "balance": 15550.33 },
      { "date": "2024-11-12", "description": "Grocery Store", "amount": -187.43, "balance": 15362.90 },
      { "date": "2024-11-18", "description": "Utility Bill", "amount": -245.73, "balance": 15117.17 }
    ]
  },
  "model": "max",
  "requestId": 44
}

Academic Search

cURL

curl -X POST https://global.pdfvector.com/api/academic/search \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "transformer attention mechanism",
    "providers": ["semantic-scholar", "pubmed", "openalex", "europe-pmc"],
    "limit": 10,
    "offset": 0,
    "yearFrom": 2017,
    "yearTo": 2025,
    "fields": ["title", "authors", "year", "doi", "citationCount"]
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.academic.search({
  query: "transformer attention mechanism",
  providers: ["semantic-scholar", "pubmed", "openalex", "europe-pmc"],
  limit: 10,
  offset: 0,
  yearFrom: 2017,
  yearTo: 2025,
  fields: ["title", "authors", "year", "doi", "totalCitations"],
});

console.log(`Found ${result.estimatedTotalResults} results`);
for (const paper of result.results) {
  console.log(`${paper.title} (${paper.year})`);
}

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/academic/search",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "query": "transformer attention mechanism",
        "providers": ["semantic-scholar", "pubmed", "openalex", "europe-pmc"],
        "limit": 10,
        "offset": 0,
        "yearFrom": 2017,
        "yearTo": 2025,
        "fields": ["title", "authors", "year", "doi", "citationCount"],
    },
)
data = response.json()
print(f"Found {data['estimatedTotalResults']} results")

Example Response

{
  "estimatedTotalResults": 48293,
  "results": [
    {
      "doi": "10.48550/arXiv.1706.03762",
      "title": "Attention Is All You Need",
      "authors": [{"name": "Ashish Vaswani"}, {"name": "Noam Shazeer"}],
      "year": 2017,
      "totalCitations": 120843,
      "abstract": "The dominant sequence transduction models...",
      "provider": "semantic-scholar"
    },
    {
      "doi": "10.18653/v1/N19-1423",
      "title": "BERT: Pre-training of Deep Bidirectional Transformers",
      "authors": [{"name": "Jacob Devlin"}, {"name": "Ming-Wei Chang"}],
      "year": 2019,
      "totalCitations": 95421,
      "provider": "semantic-scholar"
    }
  ],
  "requestId": 45
}

Academic Fetch

cURL

curl -X POST https://global.pdfvector.com/api/academic/fetch \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "ids": ["10.1038/s41586-021-03819-2", "2303.08774", "PMID:33116299"],
    "fields": ["title", "authors", "year", "doi", "abstract", "citationCount"]
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.academic.fetch({
  ids: ["10.1038/s41586-021-03819-2", "2303.08774", "PMID:33116299"],
  fields: ["title", "authors", "year", "doi", "abstract", "totalCitations"],
});

for (const paper of result.results) {
  console.log(`${paper.title} — ${paper.detectedProvider}`);
}

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/academic/fetch",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "ids": ["10.1038/s41586-021-03819-2", "2303.08774", "PMID:33116299"],
        "fields": ["title", "authors", "year", "doi", "abstract", "citationCount"],
    },
)
for paper in response.json()["results"]:
    print(paper["title"])

Example Response

{
  "results": [
    {
      "id": "10.1038/s41586-021-03819-2",
      "detectedProvider": "semantic-scholar",
      "doi": "10.1038/s41586-021-03819-2",
      "title": "Highly accurate protein structure prediction with AlphaFold",
      "authors": [{"name": "John Jumper"}, {"name": "Richard Evans"}],
      "year": 2021,
      "totalCitations": 28451,
      "url": "https://global.pdfvector.com/publications/doi:10.1038%2Fs41586-021-03819-2",
      "providerURL": "https://www.semanticscholar.org/paper/fcc0b70984..."
    },
    {
      "id": "2303.08774",
      "detectedProvider": "arxiv",
      "title": "GPT-4 Technical Report",
      "authors": [{"name": "OpenAI"}],
      "year": 2023,
      "url": "https://global.pdfvector.com/publications/arxiv:2303.08774",
      "providerURL": "https://arxiv.org/abs/2303.08774"
    }
  ],
  "requestId": 46
}

Academic Parse

cURL

curl -X POST https://global.pdfvector.com/api/academic/parse \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "id": "1706.03762",
    "model": "auto",
    "includePages": true
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.academic.parse({
  id: "1706.03762",
  model: "auto",
  includePages: true,
});

console.log(result.title);
console.log(result.pdfURL);
console.log(result.markdown);
console.log(result.pages?.[0]?.markdown);

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/academic/parse",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "url": "https://arxiv.org/abs/1706.03762",
        "model": "auto",
        "includePages": True,
    },
)
data = response.json()
print(data["title"])
print(data["markdown"])
print(data["pages"][0]["markdown"])

Example Response

{
  "id": "1706.03762",
  "title": "Attention Is All You Need",
  "doi": "10.48550/arXiv.1706.03762",
  "url": "https://global.pdfvector.com/publications/arxiv:1706.03762",
  "providerURL": "https://arxiv.org/abs/1706.03762",
  "pdfURL": "https://arxiv.org/pdf/1706.03762",
  "detectedProvider": "arxiv",
  "markdown": "# Attention Is All You Need\n\nAshish Vaswani...",
  "pageCount": 15,
  "pages": [{ "pageNumber": 1, "markdown": "# Attention Is All You Need\n\nAshish Vaswani..." }],
  "model": "auto",
  "credits": 15,
  "requestId": 47
}

Academic Find Citations

cURL

curl -X POST https://global.pdfvector.com/api/academic/findCitations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "paragraph": "Large language models have demonstrated remarkable few-shot learning capabilities across a wide range of NLP tasks. The scaling of transformer architectures to billions of parameters has led to emergent abilities not observed in smaller models.",
    "providers": ["semantic-scholar", "openalex"],
    "fields": ["title", "authors", "year", "doi", "citationCount"]
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.academic.findCitations({
  paragraph: "Large language models have demonstrated remarkable few-shot learning capabilities across a wide range of NLP tasks. The scaling of transformer architectures to billions of parameters has led to emergent abilities not observed in smaller models.",
  providers: ["semantic-scholar", "openalex"],
  fields: ["title", "authors", "year", "doi", "totalCitations"],
});

console.log(`Found ${result.totalCitations} citations`);
for (const item of result.results) {
  console.log(`"${item.sentence}"`);
  for (const cite of item.citations) {
    console.log(`  [${cite.score}] ${cite.title}`);
  }
}

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/academic/findCitations",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "paragraph": "Large language models have demonstrated remarkable few-shot learning capabilities across a wide range of NLP tasks. The scaling of transformer architectures to billions of parameters has led to emergent abilities not observed in smaller models.",
        "providers": ["semantic-scholar", "openalex"],
        "fields": ["title", "authors", "year", "doi", "citationCount"],
    },
)
data = response.json()
print(f"Found {data['totalCitations']} citations")

Example Response

{
  "results": [
    {
      "sentence": "Deep learning has revolutionized NLP.",
      "citations": [
        {
          "score": 9,
          "doi": "10.1038/nature14539",
          "title": "Deep Learning",
          "authors": [{"name": "Ian Goodfellow"}, {"name": "Yoshua Bengio"}],
          "year": 2016,
          "totalCitations": 75000
        }
      ]
    },
    {
      "sentence": "Transformer architectures have become the dominant approach for sequence modeling.",
      "citations": [
        {
          "score": 10,
          "doi": "10.48550/arXiv.1706.03762",
          "title": "Attention Is All You Need",
          "authors": [{"name": "Ashish Vaswani"}],
          "year": 2017,
          "totalCitations": 120843
        }
      ]
    }
  ],
  "sentenceCount": 2,
  "totalCitations": 2,
  "requestId": 47
}

Academic Paper Graph

cURL

curl -X POST https://global.pdfvector.com/api/academic/paperGraph \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "id": "10.1038/nature12373",
    "citationsLimit": 20,
    "referencesLimit": 20,
    "fields": ["title", "authors", "year", "doi", "totalCitations"]
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.academic.paperGraph({
  id: "10.1038/nature12373",
  citationsLimit: 20,
  referencesLimit: 20,
  fields: ["title", "authors", "year", "doi", "totalCitations"],
});

console.log(`Paper: ${result.paper.title}`);
console.log(`Citations: ${result.citations.length} of ${result.totalCitations}`);
console.log(`References: ${result.references.length} of ${result.totalReferences}`);

for (const cite of result.citations) {
  console.log(`  Cited by: ${cite.title} (${cite.year})`);
}

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/academic/paperGraph",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "id": "10.1038/nature12373",
        "citationsLimit": 20,
        "referencesLimit": 20,
        "fields": ["title", "authors", "year", "doi", "totalCitations"],
    },
)
data = response.json()
print(f"Paper: {data['paper']['title']}")
print(f"Citations: {len(data['citations'])} of {data['totalCitations']}")

Example Response

{
  "paper": {
    "doi": "10.1038/nature12373",
    "title": "Sequence-to-sequence models for data-to-text generation",
    "authors": [{"name": "Sam Wiseman"}, {"name": "Stuart Shieber"}],
    "year": 2017,
    "totalCitations": 842
  },
  "citations": [
    {
      "doi": "10.18653/v1/P19-1602",
      "title": "Data-to-Text Generation with Content Selection and Planning",
      "authors": [{"name": "Ratish Puduppully"}],
      "year": 2019,
      "totalCitations": 312
    },
    {
      "doi": "10.18653/v1/2020.acl-main.17",
      "title": "Bridging the Structural Gap Between Encoding and Decoding",
      "authors": [{"name": "Kaiqiang Song"}],
      "year": 2020,
      "totalCitations": 45
    }
  ],
  "references": [
    {
      "doi": "10.48550/arXiv.1409.3215",
      "title": "Sequence to Sequence Learning with Neural Networks",
      "authors": [{"name": "Ilya Sutskever"}],
      "year": 2014,
      "totalCitations": 25891
    }
  ],
  "totalCitations": 842,
  "totalReferences": 35,
  "requestId": 48
}

Academic Similar Papers

cURL

curl -X POST https://global.pdfvector.com/api/academic/similarPapers \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "id": "10.1038/nature12373",
    "limit": 10,
    "includeEdges": false,
    "fields": ["title", "authors", "year", "doi", "totalCitations"]
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.academic.similarPapers({
  id: "10.1038/nature12373",
  limit: 10,
  includeEdges: false,
  fields: ["title", "authors", "year", "doi", "totalCitations"],
});

console.log(`Seed: ${result.seed.title}`);
console.log(`Found ${result.results.length} similar papers`);

for (const item of result.results) {
  console.log(`  [${item.similarity.toFixed(2)}] ${item.publication.title}`);
}

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/academic/similarPapers",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "id": "10.1038/nature12373",
        "limit": 10,
        "includeEdges": False,
        "fields": ["title", "authors", "year", "doi", "totalCitations"],
    },
)
data = response.json()
print(f"Seed: {data['seed']['title']}")
for item in data["results"]:
    print(f"  [{item['similarity']:.2f}] {item['publication']['title']}")

Example Response

{
  "seed": {
    "doi": "10.1038/nature12373",
    "title": "Sequence-to-sequence models for data-to-text generation",
    "authors": [{"name": "Sam Wiseman"}],
    "year": 2017,
    "totalCitations": 842
  },
  "results": [
    {
      "publication": {
        "doi": "10.18653/v1/P19-1602",
        "title": "Data-to-Text Generation with Content Selection and Planning",
        "authors": [{"name": "Ratish Puduppully"}],
        "year": 2019,
        "totalCitations": 312
      },
      "similarity": 0.94,
      "pageRank": 0.0023,
      "distance": 1
    },
    {
      "publication": {
        "doi": "10.1162/tacl_a_00302",
        "title": "Neural Data-to-Text Generation: A Comparison between Pipeline and End-to-End Architectures",
        "authors": [{"name": "Thiago Castro Ferreira"}],
        "year": 2019,
        "totalCitations": 89
      },
      "similarity": 0.87,
      "pageRank": 0.0018,
      "distance": 1
    }
  ],
  "requestId": 49
}

Academic Search Grants

cURL

curl -X POST https://global.pdfvector.com/api/academic/searchGrants \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "machine learning healthcare",
    "providers": ["grants-gov", "nih-reporter"],
    "limit": 5,
    "fundingMin": 50000,
    "deadlineFrom": "2026-01-01",
    "fields": ["title", "agency", "fundingAmountMin", "fundingAmountMax", "deadlineDate", "url"]
  }'

TypeScript

import { createClient } from "@pdfvector/client";

const client = createClient({ apiKey: "YOUR_API_KEY" });

const result = await client.academic.searchGrants({
  query: "machine learning healthcare",
  providers: ["grants-gov", "nih-reporter"],
  limit: 5,
  fundingMin: 50000,
  deadlineFrom: "2026-01-01",
  fields: ["title", "agency", "fundingAmountMin", "fundingAmountMax", "deadlineDate", "url"],
});

console.log(`Found ${result.estimatedTotalResults} grants`);
for (const grant of result.results) {
  console.log(`${grant.title} — ${grant.agency}`);
  console.log(`  Funding: $${grant.fundingAmountMin?.toLocaleString()} - $${grant.fundingAmountMax?.toLocaleString()}`);
  console.log(`  Deadline: ${grant.deadlineDate}`);
}

Python

import requests

response = requests.post(
    "https://global.pdfvector.com/api/academic/searchGrants",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
        "query": "machine learning healthcare",
        "providers": ["grants-gov", "nih-reporter"],
        "limit": 5,
        "fundingMin": 50000,
        "deadlineFrom": "2026-01-01",
        "fields": ["title", "agency", "fundingAmountMin", "fundingAmountMax", "deadlineDate", "url"],
    },
)
data = response.json()
print(f"Found {data['estimatedTotalResults']} grants")
for grant in data["results"]:
    print(f"{grant['title']} — {grant['agency']}")

Example Response

{
  "estimatedTotalResults": 127,
  "results": [
    {
      "title": "Artificial Intelligence for Transforming Healthcare Delivery",
      "agency": "National Institutes of Health",
      "fundingAmountMin": 250000,
      "fundingAmountMax": 500000,
      "deadlineDate": "2026-06-15",
      "url": "https://grants.nih.gov/grants/guide/rfa-files/RFA-HL-25-001.html",
      "provider": "nih-reporter"
    },
    {
      "title": "Machine Learning Applications in Clinical Decision Support",
      "agency": "Department of Health and Human Services",
      "fundingAmountMin": 100000,
      "fundingAmountMax": 300000,
      "deadlineDate": "2026-03-31",
      "url": "https://www.grants.gov/search-results-detail/123456",
      "provider": "grants-gov"
    }
  ],
  "requestId": 50
}

Common Pitfalls

Do not use a real API key in generated code samples.
Do not assume pages exists unless includePages was requested.
Do not rely only on pages when the caller needs the whole document; use top-level markdown.
For Academic Parse, do not invent metadata. Use returned fields such as title, doi, providerURL, and pdfURL.

About

PDF Vector API

Sources

Authentication

TypeScript SDK

Page-Separated Output

Credit Costs

Model Guidance

Agent Workflow

Public API Examples

Document Parse

cURL

TypeScript

Python

Example Response

Document Ask

cURL

TypeScript

Python

Example Response

Document Extract

cURL

TypeScript

Python

Example Response

Identity Parse

cURL

TypeScript

Python

Example Response

Identity Ask

cURL

TypeScript

Python

Example Response

Identity Extract

cURL

TypeScript

Python

Example Response

Invoice Parse

cURL

TypeScript

Python

Example Response

Invoice Ask

cURL

TypeScript

Python

Example Response

Invoice Extract

cURL

TypeScript

Python

Example Response

Bank Statement Parse

cURL

TypeScript

Python

Example Response

Bank Statement Ask

cURL

TypeScript

Python

Example Response

Bank Statement Extract

cURL

TypeScript

Python

Example Response

Academic Search

cURL

TypeScript

Python

Example Response

Academic Fetch

cURL

TypeScript

Python

Example Response