Landing AI prices its Agentic Document Extraction in credits at $0.01 each, so nobody quotes dollars per page. Converted, the newer v2 parse is $5 to $10 per 1,000 pages, the original DPT-2 parse is $30, and extraction is billed by character. Here is every operation in dollars, plus the v2 price cut and the classify 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.
Landing AI bills its Agentic Document Extraction in credits at $0.01 each. The newer v2 parse is 0.5 to 1 credit per page ($5 to $10 per 1,000 pages), the original DPT-2 parse is 3 credits ($30), and the lighter DPT-2 mini is 1.5 credits ($15). Classify is 0.5 credits ($5), and extraction is billed by character rather than by page, so its per-1,000-pages cost is an average. The v2 generation cut the parse price 3x to 6x and made zero data retention free, where DPT-2 charged a 33 percent surcharge for it. Every account gets 1,000 free credits, but they expire 90 days after signup, unlike LlamaParse's monthly renewal. Against plain cloud OCR at $1.50 per 1,000 pages, even the cheapest Landing AI tier is about 3 times more, because you are paying for an agentic reasoning pass, not cheaper OCR.
The credit column is Landing AI's own published rate. The dollar columns convert it at $0.01 per credit.
| Operation | Credits | Per page | Per 1,000 pages | What it is |
|---|---|---|---|---|
| Parse, DPT-2 (v1) | 3 credits / page | $0.030 | $30.00 | Original agentic parse model |
| Parse, DPT-2 mini (v1) | 1.5 credits / page | $0.015 | $15.00 | Lighter v1 parse model |
| Parse v2, real-time | 1 credit / page | $0.010 | $10.00 | Newer v2 model, synchronous |
| Parse v2, async standard | 0.5 credit / page | $0.005 | $5.00 | Cheapest, batch job |
| Parse v2, output add-on | 0.5 cr / 1,000 output chars | +$0.005 | +$5.00 / 1M chars | Meters the markdown returned |
| Classify | 0.5 credit / page | $0.005 | $5.00 | Document type only, not parsing |
| Extract, v1 | 1cr/5,000 in + 1cr/1,000 out chars | by character | varies | Structured fields, character-billed |
| Extract v2, real-time | 1cr/2,500 in + 1cr/500 out chars | by character | varies | Higher output-char rate |
| Extract v2, async standard | 1cr/5,000 in + 1cr/1,000 out chars | by character | varies | Cheaper batch extraction |
| Split, v1 | 1 cr / 5,000 input chars | by character | varies | Document boundary detection |
| Zero data retention, v2 | no extra credits | $0.000 | $0.00 | Free on the v2 generation |
| Zero data retention, DPT-2 | +1 credit / page | +$0.010 | +$10.00 | 33% surcharge on DPT-2 parse |
Rates read from the Landing AI pricing and credit-consumption documentation on July 22, 2026. A credit is $0.01 ($1 buys 100). Free tier is 1,000 credits, expiring 90 days after signup. Re-verify before quoting.
The headline rates are honest, but two rules in the credit-consumption docs move the number people quote.
The number that circulates online for Landing AI, "credits equal pages times 0.5", is the Classify endpoint, not parsing. Classify is 0.5 credits per page, which is $5 per 1,000 pages. A real DPT-2 document parse is 3 credits per page, which is $30 per 1,000 pages, six times the classify figure. The v2 parse narrows the gap to $5 to $10, but the point stands: classify is not the parse rate, and budgeting on 0.5 credits per page underprices most jobs.
The v2 parse is 1 credit per page, but it adds 0.5 credits per 1,000 output characters, a meter DPT-2's flat per-page parse did not have. A text-dense page that returns 4,000 characters of markdown adds 2 credits, doubling the $0.01 headline to $0.02. So the real per-page cost of the v2 parse depends on how much text your documents emit. Sparse forms stay near the headline; dense reports and tables cost more.
Landing AI's newer v2 parse endpoints are far cheaper than the original DPT-2, and they dropped the zero-data-retention surcharge.
$30 → $10
DPT-2 parse was 3 credits per page ($30 per 1,000 pages). The v2 real-time parse is 1 credit ($10). Same synchronous call, a third of the price.
$30 → $5
The v2 async standard job is 0.5 credits per page ($5 per 1,000 pages), the cheapest agentic parse published, under LlamaParse Agentic at $12.50 and Reducto at $30 to $60.
+$10 → $0
On DPT-2, ZDR added 1 credit per page, a 33 percent surcharge. On v2 it consumes no extra credits. For regulated buyers who need ZDR, the v2 move removes a real cost.
On the v2 async parse, Landing AI is now the cheapest agentic option; on DPT-2 it was the most expensive. See the full breakdown on the agentic document extraction pricing comparison, or the head-to-head on Landing AI versus Reducto.
| Tool | Per 1,000 pages | Output | Free credits |
|---|---|---|---|
| Landing AI v2 parse, async | $5.00 | Markdown, agentic, batch | 1,000, expire in 90 days |
| Landing AI v2 parse, real-time | $10.00 | Markdown, agentic | 1,000, expire in 90 days |
| Landing AI DPT-2 parse | $30.00 | Markdown, agentic | 1,000, expire in 90 days |
| LlamaParse Fast | $1.25 | Spatial text only, no markdown | 10,000 / month, renews |
| LlamaParse Agentic | $12.50 | Markdown, tables, reasoning | 10,000 / month, renews |
| Reducto standard parse | $15.00 | Markdown, agentic optional | 15,000, one time |
| 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.
Landing AI vs Reducto
The head-to-head on price and fit
Reducto pricing
Every operation in dollars per 1,000 pages
LlamaParse pricing
Every tier in dollars per 1,000 pages
Agentic document extraction pricing
LlamaParse, Reducto and Landing AI, normalized
OCR pricing per 1,000 pages
Every vendor, one rate table
PDF to JSON API
Turn documents into structured JSON
From the same family of tools
