Open API PricingOpen API Pricing

Azure AI Document Intelligence vs Mindee pricing

Both are in document ai & ocr apis. Azure AI Document Intelligence is billed per 1k pages; Mindee is billed credit subscription tiers. Numbers below are copied from each vendor's pricing page.

Azure AI Document Intelligence vs Mindee pricing side by side
Free tier
500 pages/mo (F0, first 2 pages per request)
Billed
Per 1k pages
Best for
Prebuilt + custom models on Azure
Verified
Jun 27, 2026
MindeeFree tier or trial
Free tier
14-day free trial, no card
Billed
Credit subscription tiers
Best for
Prebuilt document parsing API
Verified
Jun 27, 2026

Azure AI Document Intelligence plans

PlanPriceWhat it covers
Free (F0)$0500 pages/month; only the first 2 pages of any analyze request processed; no SLA.
Read (OCR)$1.50 / 1,000 pagesText + handwriting OCR; drops to ~$0.60 / 1,000 pages beyond 1M pages.
Layout$10 / 1,000 pagesText, tables, selection marks, titles, headings, and document structure.
Prebuilt models$10 / 1,000 pagesInvoice, receipt, ID, W-2, health insurance card, business card, etc.
Custom extraction$30 / 1,000 pagesModels trained on your own documents; commitment tiers lower the per-page rate (~$27/$24/$21).

Source: https://azure.microsoft.com/en-us/pricing/details/document-intelligence/

Mindee plans

PlanPriceWhat it covers
StarterEUR44/month (EUR528/yr)500 monthly credits, community support, unlimited models; extra credits EUR0.05 each.
ProEUR179/month (EUR2,148/yr)Most popular. 2,500 monthly credits, live chat support, team members, RAG up to 20 docs; extra credits EUR0.04.
BusinessEUR584/month (EUR7,008/yr)10,000 monthly credits, priority support, unlimited RAG, confidence scores; extra credits EUR0.035.
EnterpriseCustomFor 250k+ credits/year: dedicated account manager, custom SLAs, premium technical support, tailored pricing.
Free trialEUR0 (14 days)Time-limited trial; no permanent free plan offered.

Source: https://www.mindee.com/pricing

Azure AI Document Intelligence pricing guideMindee pricing guideQuality scores on APIbenchmarks →