# DocsFlow AI

> DocsFlow AI is an AI-powered document intelligence platform that converts PDFs, scans, invoices, IDs, and websites into structured JSON via OCR, web crawling, and automated validation APIs.

## Key facts

- **Name:** DocsFlow AI
- **Domain:** https://docsflowai.com
- **Founded:** 2026
- **Headquarters:** Hosur, Tamil Nadu, India
- **Support:** support@docsflowai.com
- **Free tier:** 100 documents/month, 10 web crawls/month (no credit card required)

## What DocsFlow AI does

- Extracts structured data (text, tables, line items, signatures, metadata) from PDFs, images, scans, invoices, contracts, IDs, and forms.
- Crawls websites — including JavaScript-rendered pages — and extracts structured data, content, metadata, and links.
- Returns machine-readable JSON via a REST API, with SDKs, webhooks, and integrations for Zapier, Make, Slack, Google Sheets, Salesforce, and more.
- Processes over 50 document formats without requiring templates.

## Pages

- [Home](https://docsflowai.com/): DocsFlow AI is an AI-powered document intelligence platform that converts PDFs, scans, invoices, contracts, IDs, and websites into structured JSON — OCR, document parsing, and web crawling APIs in one platform
- [Features](https://docsflowai.com/features): full document AI feature set — OCR for printed and handwritten text, PDF-to-JSON, complex table extraction, custom extraction schemas, signature and stamp detection, validation with confidence scoring, and 50+ integrations
- [Compare — DocsFlow AI vs Base64.ai, Nanonets & AWS Textract](https://docsflowai.com/compare): head-to-head comparisons with Reducto, LlamaParse, Unstructured, and Base64.ai on accuracy, pricing, JSON output, web crawling, data retention, and compliance
- [DocsFlow AI vs Reducto](https://docsflowai.com/compare/reducto)
- [DocsFlow AI vs LlamaParse](https://docsflowai.com/compare/llamaparse)
- [DocsFlow AI vs Unstructured](https://docsflowai.com/compare/unstructured)
- [DocsFlow AI vs Base64.ai](https://docsflowai.com/compare/base64.ai)
- [Pricing](https://docsflowai.com/pricing): Starter free forever with 100 documents/month, Pro from $99/month for 10,000 documents + 500 web crawls, Enterprise custom — no setup fees, no credit card required to start
- [Documentation & API Reference](https://docsflowai.com/docs): Quickstart, endpoints, authentication, and SDKs for the OCR, document parsing, and web crawling API
- [OCR API](https://docsflowai.com/ocr-api): extract validated JSON from PDFs, scanned images, invoices, contracts, and IDs with 99.2% accuracy — multi-model AI OCR, vision-language models, custom extraction schemas, confidence scores, 50+ formats, 100+ languages, zero-template
- [Document Parsing API](https://docsflowai.com/document-parsing-api): parse PDF, DOCX, XLSX, HTML, and scanned images into validated JSON with semantic field labels, table extraction (merged cells, multi-page), form fields, and confidence scores — 30+ formats, 100+ languages, zero-template, LLM + vision-language models
- [Web Crawling API](https://docsflowai.com/web-crawling-api): crawl any URL — static or JavaScript-rendered — and receive clean JSON, Markdown, or HTML. Headless Chromium rendering, proxy rotation across 150+ countries, bulk sitemap crawls up to 10,000 URLs, custom extraction schemas, and LLM-ready output
- [Intelligent Document Processing](https://docsflowai.com/intelligent-document-processing): AI-powered IDP platform automating extraction, classification, and validation across 30+ document formats — zero-template LLM extraction, vision-language models, validation rules, confidence scoring, real-time webhook delivery, SOC 2 · GDPR · HIPAA
- [Contract Analysis](https://docsflowai.com/contract-analysis): AI contract analysis with 99.2% clause extraction accuracy in under 30 seconds — clause extraction, risk flag detection, critical dates, obligation tracking, contract comparison, and M&A due diligence in 40+ languages
- [Healthcare Automation](https://docsflowai.com/healthcare-automation): HIPAA-compliant AI medical document processing — prescription OCR (handwritten), lab report extraction, discharge summaries, insurance forms, ICD-10/CPT coding, and HL7 FHIR R4 export in under 4 seconds
- [KYC & Onboarding](https://docsflowai.com/kyc-onboarding): AI identity verification from 180+ countries in under 3 seconds — passports, national IDs, driver's licences, proof of address, with MRZ cross-checks, liveness detection, and OFAC/UN/EU/HM Treasury sanctions screening
- [Invoice & Payments API](https://docsflowai.com/invoice-payments-api): AI invoice data extraction with 99.4% accuracy in under 2 seconds, PO matching, validation, ERP integration (SAP, Oracle, NetSuite, QuickBooks, Xero), and AP automation
- [Integrations](https://docsflowai.com/integrations): Zapier, Make, Salesforce, HubSpot, Snowflake, QuickBooks, Slack, Google Sheets, and 50+ native connectors via REST API and webhooks
- [Benchmarks](https://docsflowai.com/benchmark): 99.2% average extraction accuracy, <1s median single-doc latency, 10,000+ docs/min batch throughput, measured on 50,000+ documents
- [Security & Compliance](https://docsflowai.com/security): SOC 2 Type II, HIPAA, GDPR, ISO 27001, AES-256 encryption, zero data retention, per-request isolated sandboxes, immutable audit logs
- [About](https://docsflowai.com/about): AI-powered document intelligence platform founded in 2026, headquartered in Hosur, Tamil Nadu, India
- [Contact](https://docsflowai.com/contact): sales, support, and enterprise inquiries — support@docsflowai.com, +91 8300365605
- [Blog](https://docsflowai.com/blog): tutorials and deep dives on document processing, OCR, web crawling, and intelligent document data extraction
- [Privacy Policy](https://docsflowai.com/privacy)
- [Terms of Service](https://docsflowai.com/terms)

## Frequently asked questions

- **What is DocsFlow AI?** DocsFlow AI is a document intelligence platform that extracts, understands, and transforms data from PDFs, forms, invoices, contracts, images, and websites into structured, machine-readable JSON using OCR, web crawling, and AI-powered extraction.
- **How accurate is DocsFlow AI?** Up to 99.2% extraction accuracy using advanced OCR, multi-model AI pipelines, and intelligent field detection.
- **Is DocsFlow AI secure?** DocsFlow AI uses GDPR and SOC 2-aligned security, isolated processing environments, audit logging, and zero data retention after processing.
- **Is DocsFlow AI SOC 2 certified?** Yes. DocsFlow AI holds SOC 2 Type II certification, audited annually by an independent third party. The audit covers security, availability, processing integrity, and confidentiality controls. Audit reports are available on request.
- **Does DocsFlow AI store documents after processing?** No. DocsFlow AI enforces a zero data retention policy by default — files are cryptographically purged from all systems within seconds of extraction completing, with no backups, caching, or content logs.
- **How does DocsFlow AI encrypt documents?** Documents are encrypted in transit using TLS 1.3 and at rest using AES-256. Keys are unique per request, rotated on a 90-day schedule, and destroyed immediately after extraction.
- **Is DocsFlow AI HIPAA compliant for processing patient records?** Yes. DocsFlow AI is HIPAA compliant with AES-256 encryption at rest, TLS 1.3 in transit, isolated ephemeral sandboxes per document, zero data retention mode that purges source documents immediately after extraction, and BAAs available for covered entities on Enterprise plans.
- **Can DocsFlow AI read handwritten prescriptions?** Yes. DocsFlow AI uses a specialized handwriting OCR model trained on clinical documents to extract drug names, dosages, frequencies, prescriber information, and patient identifiers — including physician shorthand and abbreviations. Medication names can be normalized to RxCUI or SNOMED CT identifiers.
- **Does DocsFlow AI support HL7 FHIR export?** Yes. DocsFlow AI supports HL7 FHIR R4 output. Extracted clinical data maps to FHIR resources including Patient, Observation, DiagnosticReport, MedicationRequest, and Encounter, and can be pushed directly to any FHIR-compliant EHR via REST API webhook.
- **What healthcare document types does DocsFlow AI support?** DocsFlow AI processes handwritten prescriptions, structured lab reports, discharge summaries, insurance prior authorization forms, clinical trial documents, provider credentialing packets, referral letters, and operative reports — typed and handwritten, in 40+ languages, with automatic document type detection.
- **Can DocsFlow AI extract ICD-10 and CPT codes automatically?** Yes. DocsFlow AI's clinical NLP model automatically maps diagnoses to ICD-10-CM codes and procedures to CPT codes, with optional SNOMED CT concepts. This eliminates manual medical coding for structured document types and accelerates reimbursement cycles.
- **How does DocsFlow AI handle prior authorization automation?** DocsFlow AI extracts policy numbers, member IDs, diagnosis codes, procedure codes, and clinical justification text from prior authorization documents, then routes the structured data to payer portals or internal approval workflows — reducing prior-auth processing from days to hours.
- **What is the DocsFlow AI API?** The DocsFlow AI API is a REST interface that converts PDFs, scans, invoices, IDs, contracts, and websites into structured JSON with a single HTTP call. Responses include validated fields, tables, line items, and per-field confidence scores.
- **Is DocsFlow AI really free to start?** Yes. The Starter plan is free forever with 100 documents per month and 10 web crawls per month. No credit card is required to create an account.
- **How much does DocsFlow AI cost?** Starter is free forever (100 documents/month). Pro is $99/month for 10,000 documents and 500 web crawls per month. Enterprise is custom-priced with SSO/SAML, custom AI models, on-premise options, and SLA guarantees. No setup fees.
- **What counts as one document?** One document is a single file — PDF, image, scan, Word, or PowerPoint — processed through the extraction API. A web crawl is one page fetched and parsed by the web crawling API. Pricing is per document, not per page.
- **Are there any setup or overage fees?** No setup fees. If you exceed your monthly limits, you can upgrade to a higher plan or buy additional usage — there are no surprise overage charges. Plans can be upgraded, downgraded, or cancelled at any time.
- **Does DocsFlow AI offer an annual discount?** Yes — annual plans are available at a discount. Select the annual option in your account or contact support@docsflowai.com to see the rate.
- **Which endpoints does the DocsFlow AI API provide?** Five core endpoints: POST /v1/parse (extract structured data from documents), POST /v1/crawl (async website crawl including JavaScript-rendered pages), POST /v1/parse/url (parse from a download URL), GET /v1/jobs/:id (poll job status), and GET /v1/account (plan limits and usage).
- **How do I authenticate to the DocsFlow AI API?** Pass your API key as an Authorization: Bearer header on every request using HTTPS with TLS 1.2+. Keys are scoped to read-only, write, or admin permissions and can be rotated from the dashboard without downtime.
- **Which SDKs are available for DocsFlow AI?** Official SDKs cover Python (3.8+), Node.js (ESM and CommonJS), Go (1.20+), and Ruby (3.0+), plus a CLI, an OpenAPI 3.0 spec, webhooks with retry, and a streaming NDJSON endpoint.
- **What is AI invoice data extraction?** AI invoice data extraction uses computer vision and NLP to read structured fields — invoice number, vendor, line items, totals, payment terms — from any invoice and return them as JSON. It works without vendor templates, generalising across layouts automatically.
- **What invoice fields can DocsFlow AI extract?** DocsFlow AI extracts invoice number, vendor and buyer details, VAT/tax, issue and due dates, payment terms, line items (description, quantity, unit price, subtotal), totals, bank details (IBAN, SWIFT), PO references, and currency, plus any custom schema fields.
- **How accurate is DocsFlow AI invoice extraction?** DocsFlow AI achieves 99.4% extraction accuracy with cross-field validation: line-item sums are verified against declared totals, tax against line-level rates, and PO numbers against your reference list. Per-field confidence scores route exceptions to review.
- **Which ERPs does DocsFlow AI integrate with?** DocsFlow AI integrates with SAP, Oracle, NetSuite, Microsoft Dynamics 365, QuickBooks, Xero, Sage, Workday, Coupa, and Ariba through REST API, webhooks, and native connectors.
- **How fast is invoice extraction?** DocsFlow AI returns results in under 2 seconds on average for single invoices. Batch jobs support up to 10,000 invoices per submission streamed as NDJSON, with P99 response time under 3 seconds.
- **What identity documents can DocsFlow AI verify?** DocsFlow AI supports passports, national ID cards, driver's licences, proof of address documents (utility bills, bank statements), business registration documents, tax forms, visas, and residence permits. Document coverage spans 180+ countries with automatic document type detection — no configuration required.
- **Does DocsFlow AI include AML and sanctions screening?** Yes. Every verification call automatically screens extracted names against OFAC, UN, EU, and HM Treasury consolidated sanctions lists, as well as PEP (Politically Exposed Persons) databases covering 180+ countries. Results are returned in the same JSON response with pass/fail status per check.
- **How fast is the KYC verification decision?** DocsFlow AI returns a full verification decision — document extraction, field validation, MRZ cross-check, and sanctions screening — in under 3 seconds on average, with a P99 under 3 seconds and a 99.9% uptime SLA. Decisions are delivered via signed webhook push or polled via REST.
- **Is DocsFlow AI KYC verification GDPR compliant?** Yes. DocsFlow AI is GDPR compliant with data residency controls (EU, US, UK, APAC), a zero data retention mode that purges all source documents and intermediate data immediately after extraction, and immutable audit logs for regulatory review.
- **How does DocsFlow AI integrate with existing KYC workflows?** DocsFlow AI provides a REST API with SDKs for Python, Node.js, Go, Ruby, and PHP. Verification results are delivered via webhook push with HMAC signature verification, or polled via REST. The structured JSON includes all extracted fields, per-field confidence scores, and per-check pass/fail results. Batch requests support up to 500 verifications per request.
- **What is KYC document verification?** KYC (Know Your Customer) document verification confirms a person's identity by extracting and validating information from government-issued identity documents. Regulated entities — banks, fintechs, crypto exchanges, and HR departments — are required to verify customer identities before onboarding. AI-powered verification automates this with OCR, machine learning, and sanctions screening in a single API call returning a structured decision in under 3 seconds.
- **What types of contracts can DocsFlow AI analyze?** DocsFlow AI analyzes any contract type — NDAs, MSAs, SaaS agreements, employment contracts, vendor agreements, lease agreements, M&A transaction documents, and healthcare BAAs. It supports PDF, DOCX, scanned images, and digitally signed documents in 40+ languages.
- **How accurate is AI contract clause extraction?** DocsFlow AI achieves up to 99.2% clause extraction accuracy using a multi-model AI pipeline that understands document context, not just keywords. Confidence scores are returned per field so low-confidence extractions can be flagged for human review.
- **Can DocsFlow AI detect risky or unusual clauses?** Yes. DocsFlow AI automatically flags one-sided clauses, uncapped liabilities, missing standard protections, and unusual termination rights. Each flag comes with a plain-language explanation so non-lawyers can understand the risk without reading the original clause.
- **How does contract comparison work?** Upload two versions of a contract and DocsFlow AI returns a structured diff of every changed clause — what was added, removed, or softened — along with a risk delta score showing whether the new version is more or less favorable than the original.
- **Is contract data kept secure and confidential?** Yes. Every contract is processed in an isolated, ephemeral sandbox with AES-256 encryption at rest and TLS 1.3 in transit. Zero data retention is available by default — documents are purged immediately after extraction. DocsFlow AI is SOC 2, GDPR, and HIPAA compliant.
- **Can I analyze my entire existing contract portfolio in bulk?** Yes. DocsFlow AI supports bulk contract processing — upload thousands of existing contracts via API or direct upload and have them all analyzed in parallel, commonly used to retroactively build a searchable contract intelligence repository.
- **What is a web crawling API?** A web crawling API accepts URLs over HTTP and returns the page content — as HTML, Markdown, or structured JSON — without requiring the caller to manage browser infrastructure, proxy pools, or anti-bot bypass logic. DocsFlow AI handles all of that server-side.
- **Does the web crawling API support JavaScript-rendered pages?** Yes. Set render_js: true in the request body and DocsFlow AI runs headless Chromium, executes all client-side JavaScript, and waits for DOM stability before extracting content. React, Next.js, Vue, and Angular SPAs are all supported.
- **How does bulk crawling work?** POST to /v1/crawl/bulk with up to 10,000 URLs — or a sitemap URL. Results are delivered to your webhook endpoint as each page completes. Configurable concurrency lets you balance speed against target-server politeness.
- **Do I need to manage proxies?** No. DocsFlow AI manages a residential and datacenter proxy pool across 150+ countries with automatic rotation. No proxy procurement, configuration, or rotation logic is required on your end.
- **What output formats does the web crawling API return?** The API returns content in three formats: HTML (cleaned rendered HTML), Markdown (structured Markdown ready for LLM ingestion), and JSON (structured fields extracted according to your output_schema). Set output_format in the request body to choose.
- **Is the web crawling API suitable for AI training data collection?** Yes. The Markdown output mode returns clean, boilerplate-free content with semantic structure preserved — headings, paragraphs, code blocks, and tables. This format is ready for direct ingestion into pre-training pipelines, fine-tuning datasets, and RAG vector databases without additional preprocessing.
- **What is an OCR API?** An OCR API accepts document or image files over HTTP and returns machine-readable text or structured data. The DocsFlow AI OCR API goes beyond raw text — it understands document layout, detects business fields and tables, and returns validated JSON ready for automation pipelines.
- **How accurate is the DocsFlow AI OCR API?** DocsFlow AI achieves up to 99.2% average extraction accuracy using a multi-model AI pipeline: an OCR layer, a layout analysis model, and a semantic field mapper. Cross-model agreement and a validation layer produce per-field confidence scores before results reach your application.
- **Which file formats does the OCR API support?** The DocsFlow AI OCR API supports over 50 file formats including PDF (digital and scanned), PNG, JPG, TIFF, WEBP, BMP, HEIC, DOCX, and XLSX. It automatically detects document type, language, and page orientation without configuration.
- **Can I define custom fields for the OCR API to extract?** Yes. You can pass an output_schema in the request body that lists the exact field names and types you need. The API maps document content to your schema automatically, even when document layouts vary across vendors or formats.
- **Is the OCR API HIPAA and GDPR compliant?** Yes. DocsFlow AI provides SOC 2 Type II compliance, GDPR data processing agreements, and HIPAA-compatible zero-retention processing. Documents are processed in isolated environments and can be configured to never persist to disk.
- **How does DocsFlow AI OCR API compare to Google Cloud Vision and AWS Textract?** Google Cloud Vision and AWS Textract return raw text and bounding boxes. DocsFlow AI returns fully structured, validated JSON with labelled fields and confidence scores from a single endpoint — reducing integration effort from weeks to hours. Pricing is flat-rate with 100 free documents per month rather than unpredictable per-page billing.
- **What is document parsing?** Document parsing converts human-readable files — PDFs, Word documents, spreadsheets, scanned images — into structured, machine-readable data. A modern AI document parser like DocsFlow AI extracts semantic fields, preserves table relationships, and returns typed JSON with confidence scores — going far beyond raw text extraction.
- **Which file formats does the document parsing API support?** DocsFlow AI supports 30+ formats: PDF (digital and scanned), DOCX, XLSX, PPTX, HTML, Markdown, TXT, PNG, JPG, TIFF, WEBP, BMP, HEIC, ODS, ODT, and more. Format and language are detected automatically — no configuration required.
- **How does DocsFlow AI handle complex tables in documents?** DocsFlow AI uses a dedicated table reconstruction model that handles merged cells, multi-row headers, nested sub-tables, and tables spanning multiple pages. Tables are returned as typed two-dimensional arrays with inferred column names — ready for database insertion or spreadsheet export.
- **Can I define my own output schema for document parsing?** Yes. Pass an output_schema in the API request listing the exact field names and types you need. The AI maps document content to your schema regardless of layout variation across vendors and document versions — no re-training or template updates required.
- **Is the document parsing API HIPAA and GDPR compliant?** Yes. DocsFlow AI provides SOC 2 Type II compliance, GDPR data processing agreements, and HIPAA-compatible zero-retention processing. Documents are processed in isolated per-request environments and can be configured to never persist to disk.
- **How fast is document parsing with DocsFlow AI?** Median processing time for a typical 10-page PDF is under 3 seconds. Batch jobs scale horizontally across parallel processing pipelines, so submitting thousands of documents does not introduce per-file queue delays.
- **What is Intelligent Document Processing?** Intelligent Document Processing (IDP) is an AI-powered technology that automates the extraction, classification, validation, and routing of data from unstructured documents. Unlike traditional OCR, IDP understands document semantics — returning typed, structured JSON with confidence scores rather than raw text.
- **How is DocsFlow AI's IDP different from OCR?** OCR converts pixels to text. IDP understands meaning. DocsFlow AI combines computer vision OCR with large language models that reason about field relationships, classify document types, and extract semantically labelled values — correctly, across any vendor format, without templates.
- **Which document types does the IDP platform support?** DocsFlow AI handles invoices, purchase orders, contracts, medical records, insurance claims, shipping documents, ID documents, bank statements, tax forms, and any custom document type. All formats — PDF, DOCX, XLSX, scanned images, and 30+ more — are accepted without preprocessing.
- **Does IDP require training on my specific document templates?** No. DocsFlow AI uses zero-template extraction — pass an output schema listing the field names you need, and the AI maps document content to your schema regardless of layout variations between vendors or document versions.
- **How fast is intelligent document processing with DocsFlow AI?** Median processing time for a 10-page PDF is under 3 seconds. Batch jobs up to 10,000 documents process in parallel across horizontal pipeline workers — results stream to your webhook as each document completes.
- **Is the IDP platform compliant with HIPAA and GDPR?** Yes. Each document is processed in an isolated, ephemeral environment. Zero-retention mode ensures no document content persists after the API response. DocsFlow AI holds SOC 2 Type II certification and provides GDPR data processing agreements and HIPAA-compatible processing for healthcare workflows.
- **What does it cost?** Starter is free forever (100 documents/month). Pro is $99/month (10,000 documents + 500 crawls). Enterprise is custom-priced.
- **Does it crawl JavaScript websites?** Yes, DocsFlow AI crawls dynamic and JavaScript-rendered web pages.
- **What integrations does DocsFlow AI support?** DocsFlow AI ships 50+ native integrations plus a REST API and webhooks, covering automation (Zapier, Make, n8n, Pipedream), CRMs (Salesforce, HubSpot, Pipedrive, Zoho CRM), databases (Snowflake, BigQuery, Postgres, MySQL), finance (QuickBooks, Xero, Stripe, NetSuite), storage, communications, and spreadsheets.
- **How do DocsFlow AI webhooks work?** When extraction completes, DocsFlow AI posts a signed JSON payload to your endpoint. Payloads are verified with HMAC-SHA256 signatures, and failed deliveries retry automatically with exponential backoff up to five times.
- **How accurate is DocsFlow AI?** DocsFlow AI achieves 99.2% average extraction accuracy, measured per-field on a 50,000-document evaluation corpus. Accuracy by document type: 99.4% on invoices, 99.8% on passports and IDs, 98.7% on contracts, and 97.9% on medical records.
- **How fast is DocsFlow AI?** Median single-document latency is under 1 second for single-page PDFs; 10-page contracts complete in about 2.1 seconds at p50. Batch jobs process over 10,000 documents per minute in parallel.
- **How are DocsFlow AI benchmark numbers measured?** Accuracy is measured per-field on a 50,000-document hold-out set with ground-truth annotations. Latency is full round-trip time (upload to validated JSON) at p50/p95/p99 percentiles over 1,000 requests per document type in the US-East production region. Any team can reproduce results via the public evaluation endpoint.
- **How does DocsFlow AI compare to Base64.ai?** Both are full document intelligence platforms. DocsFlow AI differentiates with transparent public pricing, JavaScript-rendered web crawling, zero data retention by default, and a pay-per-use model with 100 free documents per month. Its 99.2% accuracy is measured on a public hold-out set rather than an internal benchmark.

## Getting started

1. Create a free account at https://alpha.docsflowai.com/login.
2. Get your API key.
3. POST to `https://api.docsflowai.com/v1/parse` with your document, or `https://api.docsflowai.com/v1/crawl` for websites.
4. Receive structured JSON with fields, tables, and confidence scores.