Extend prices in credits at $0.0125 each, so nobody quotes dollars per page. Converted, a Performance parse is $25 per 1,000 pages, extraction is $62.50 once the automatic parse is counted, and Light mode cuts it about fourfold. Here is every operation in dollars, plus the auto-parse stacking trap.
Upload a document to extract
Drop files here or click to upload
Up to 50 files
Uploading...
A credit rate says nothing about whether the fields come out right on your documents. Drop one in and check the output before you commit.
Extend bills in credits at $0.0125 each on Pay As You Go. In Performance mode a parse is 2 credits per page ($25 per 1,000 pages), and a structured extraction is 5 credits ($62.50) once you count the parse that runs automatically before it. Classification is 3 credits ($37.50) and splitting 4 credits ($50), both including that auto-parse. Light mode swaps in lighter models and cuts each operation by roughly four times, so extraction drops to about $15 per 1,000 pages. Optional review and correction agents add 1 credit ($12.50) each, and a fully agentic extraction can reach 8 credits per page ($100). Every account starts with 10,000 free credits, and the Scale plan is $500 a month for 50,000 credits at $0.01 each. Against plain cloud OCR at $1.50 per 1,000 pages, Extend is 17 to 40 times more, because you are paying for an agent-ready pipeline, not cheaper OCR.
The credit column is Extend's own published rate. The dollar columns convert it at $0.0125 per credit on Pay As You Go.
| Operation | Credits | Per page | Per 1,000 pages | What it is |
|---|---|---|---|---|
| Parse (Performance) | 2 credits / page | $0.025 | $25.00 | Layout parse to markdown, runs before every other op |
| Extract (Performance) | 3 cr + 2 cr parse = 5 | $0.0625 | $62.50 | Structured fields; the auto-parse stacks on top |
| Classify (Performance) | 1 cr + 2 cr parse = 3 | $0.0375 | $37.50 | Document type; auto-parse stacks |
| Split (Performance) | 2 cr + 2 cr parse = 4 | $0.050 | $50.00 | Boundary detection; auto-parse stacks |
| Edit Filling | 1 credit / page | $0.0125 | $12.50 | Fill a form template; no auto-parse |
| Form Detection | 2 credits / page | $0.025 | $25.00 | Detect fillable fields |
| Parse (Light) | 0.5 credit / page | $0.00625 | $6.25 | Lighter model, clean documents |
| Extract (Light) | 0.7 cr + 0.5 parse = 1.2 | $0.015 | $15.00 | Light-model extraction with auto-parse |
| Classify (Light) | 0.3 cr + 0.5 parse = 0.8 | $0.010 | $10.00 | Light-model classification |
| Split (Light) | 0.5 cr + 0.5 parse = 1 | $0.0125 | $12.50 | Light-model splitting |
| Extraction Review Agent | +1 credit / page | +$0.0125 | +$12.50 | Optional second-pass accuracy check |
| Agentic Text or Table Correction | +1 credit / page each | +$0.0125 | +$12.50 | Optional, per correction agent |
| Advanced Excel Parsing | 3 credits / 1,000 cells | by cell | varies | Spreadsheet add-on |
Rates read from the Extend pricing page and credit-consumption documentation on July 22, 2026. A credit is $0.0125 on Pay As You Go, or $0.01 on the $500-a-month Scale plan. Free tier is 10,000 credits to start. Re-verify before quoting.
The headline rates are honest, but two rules in the credit-consumption docs move the number people quote.
Extend parses a document automatically before it extracts, classifies or splits, and it bills both operations. So the 3-credit extraction is really 3 plus 2 parse credits, which is 5 credits per page, or $62.50 per 1,000 pages, not the $37.50 the 3-credit number suggests. Classification becomes 3 credits and splitting 4. If you already parsed a document, reusing that result avoids paying the parse twice, the same way Reducto lets you extract against a job id. Budget on the stacked total.
The Performance-versus-Light switch is a roughly 4x lever you set per request: a Performance extraction is 5 credits ($62.50 per 1,000 pages) and the same job in Light mode is about 1.2 credits ($15). On top of that, the optional Extraction Review Agent, Agentic Text Correction and Agentic Table Correction each add 1 credit per page ($12.50). Turn all three on with a Performance extraction and you reach 8 credits per page, which is $100 per 1,000 pages, the real ceiling.
Extend sells document infrastructure for AI agents, not a raw OCR meter, and the price reflects the extra work in the pipeline.
$1.50 → $25
Cloud OCR returns text and boxes at $1.50 per 1,000 pages. An Extend parse returns clean markdown with layout at $25, because a vision-language model rebuilds the structure rather than dumping characters.
$62.50
Performance extraction returns the fields you defined, validated, at $62.50 per 1,000 pages. That folds classification, parsing and extraction into a single agent-ready call, which is what the premium buys.
$15
Clean, born-digital documents rarely need the heavy model. Light-mode extraction at about $15 per 1,000 pages is the single biggest saving on Extend, and it is one flag per request.
Extend sits at the expensive end of agentic extraction in Performance mode and mid-pack in Light mode. See the full breakdown on the agentic document extraction pricing comparison, or the Reducto pricing reference.
| Tool | Per 1,000 pages | Output | Free credits |
|---|---|---|---|
| Extend extract (Performance) | $62.50 | Schema fields, agent-ready | 10,000, to start |
| Extend extract (Light) | $15.00 | Lighter models, clean docs | 10,000, to start |
| Extend parse (Performance) | $25.00 | Markdown | 10,000, to start |
| Reducto standard extract | $30.00 | Structured fields | 15,000, one time |
| Landing AI DPT-2 parse | $30.00 | Markdown, agentic | 1,000, expire in 90 days |
| LlamaParse Agentic | $12.50 | Markdown, tables, reasoning | 10,000 / month, renews |
| Cloud OCR (Textract, Azure, Google) | $1.50 | Raw text and key-values | Free tier varies |
A credit rate tells you nothing about accuracy. Upload a document to DocuOCR, see the extracted fields, and compare before you build a pipeline around any parser.
Reducto pricing
Every operation in dollars per 1,000 pages
Landing AI pricing
DPT-2 and the cheaper v2 parse, per 1,000 pages
LlamaParse pricing
Every tier in dollars per 1,000 pages
Agentic document extraction pricing
LlamaParse, Reducto, Landing AI and Extend, normalized
Unstructured pricing
Per-page rate and the $3,000 spend cap
OCR pricing per 1,000 pages
Every vendor, one rate table
From the same family of tools