Per-page pricing punishes long documents.

Almost every document-extraction vendor — Unstructured.io, LlamaParse, AWS Textract, Azure AI Document Intelligence, and Mindee among them — bills per page. It's a fair proxy for the OCR and inference cost of processing a page. But it has a side effect: the price of extracting a document is a function of that document's length, not of how much work it is to you — one file, fetched or uploaded, handled once.

Consider two customers sending 100 documents a month. Customer A sends one-page invoices. Customer B sends 300-page compliance reports. Under per-page billing, Customer B pays 300x more — for the same "100 documents a month" workload. Under per-document billing, both pay the same, because the unit of billing matches the unit of work: a document extracted, once.

This isn't an argument that per-page pricing is dishonest — it maps cleanly to compute cost, and for vendors selling OCR as commodity infrastructure, it's a defensible model. It's an argument that the pricing unit should match how you actually think about your workload. If you ingest long documents, per-page billing means your costs are unpredictable and tied to content you don't control (a vendor could send you a 50-page PDF instead of a 5-page one and your bill jumps 10x). txtfetch bills per document instead: a 300-page PDF and a one-page memo cost the same to extract, because from your side, they're both "one thing I needed turned into text."

The math, worked out

Below is a representative per-page rate — the Read/OCR tier that AWS Textract and Azure AI Document Intelligence both charge, and roughly what "commodity OCR" costs across the market. Try your own document length and monthly volume.

Where per-page pricing is the right call

To be fair to the vendors above: if your documents are uniformly short (single-page forms, receipts, IDs) or you need per-feature pricing so you only pay for the specific analysis you use (tables, key-value pairs, a specific prebuilt model), per-page billing can end up cheaper than a flat document quota. It's also the more transparent model for pure infrastructure — you're buying compute, not a product tier. See the individual comparisons below for exactly where each vendor's model wins for a given workload.

Read the individual comparisons

• txtfetch vs Unstructured.io — Open-core ETL for RAG pipelines, priced per page.
• txtfetch vs LlamaParse — Credit-metered parsing, tuned for complex PDFs.
• txtfetch vs AWS Textract — AWS-native OCR and document analysis, billed per 1,000 pages.
• txtfetch vs Azure AI Document Intelligence — Microsoft's prebuilt-model document API, billed per 1,000 pages.
• txtfetch vs Mindee — Subscription + credits for structured field extraction.

Full plan details, including the durable free tier, are on the pricing page.

Sources

• Unstructured.io — Pricing— accessed 2026-07
• LlamaIndex — LlamaParse pricing— accessed 2026-07
• LlamaParse free-tier credits (secondary source)— accessed 2026-07directional
• AWS Textract — Pricing— accessed 2026-07
• Azure AI Document Intelligence — Pricing— accessed 2026-07directional
• Microsoft Q&A — Document Intelligence per-page rate consensus— accessed 2026-07directional
• Mindee — Pricing— accessed 2026-07
• Mindee — derived effective per-page rate (not a published figure)— accessed 2026-07directional