Landing AI moved Agentic Document Extraction to ADE Gen2 on DPT-3 in September 2026, and the price of a page now depends on three things: the model, the service tier, and how much text the page returns. Converted at $0.01 per credit, a typical page costs $5.50 to $30 per 1,000. A synchronous call always costs twice what the same page costs as an async job.
Upload a document to extract
Drop files here or click to upload
Up to 50 files
The free demo extracts the first 5 files, the rest unlock when you upgrade
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 sells credits at $0.01 each. On ADE Gen2, every parsed page is billed as a per-page rate plus a rate per 1,000 output characters, so the price follows how much markdown a page returns. There are two models and two service tiers. DPT-3 Pro, the default, costs $30 per 1,000 typical 4,000-character pages on the priority tier and $15 on the standard tier. DPT-3 Verity, a lower-latency preview model for digitally created Latin-script documents, costs $11 on priority and $5.50 on standard. Every synchronous call runs on priority, and async jobs default to standard at half the rate. Gen2 Extract is billed by characters, zero data retention adds nothing, and Ground is free. The older Gen1 Parse is still documented at a flat 3 credits per page ($30 per 1,000 pages).
Each page costs its per-page rate plus its output-character rate. The last two columns price a typical business page that returns 4,000 characters of markdown, converted at $0.01 per credit.
| Model and tier | Applies to | Per page | Per 1,000 output chars | Typical page | Per 1,000 typical pages |
|---|---|---|---|---|---|
| DPT-3 Pro, priority | Sync calls, and jobs that request it | 1 credit | 0.5 credits | 3 credits ($0.030) | $30.00 |
| DPT-3 Pro, standard | Async jobs only, the job default | 0.5 credits | 0.25 credits | 1.5 credits ($0.015) | $15.00 |
| DPT-3 Verity, priority | Sync calls, and jobs that request it | 0.3 credits | 0.2 credits | 1.1 credits ($0.011) | $11.00 |
| DPT-3 Verity, standard | Async jobs only, the job default | 0.15 credits | 0.1 credits | 0.55 credits ($0.0055) | $5.50 |
Landing AI rounds every credit total up to the nearest 0.1 credit, so a one-page Verity standard request at 0.55 credits bills 0.6. A dense page that returns 8,000 characters costs 5 credits on DPT-3 Pro priority, which is $50 per 1,000 pages.
| Operation | Credits | In dollars | Notes |
|---|---|---|---|
| Extract, priority | 1 credit / 2,500 input chars + 1 credit / 500 output chars | by character | Sync calls, and jobs that request it |
| Extract, standard | 1 credit / 5,000 input chars + 1 credit / 1,000 output chars | about $0.015 per typical document | Median of about 1.5 credits per typical business document |
| Zero data retention | no extra credits | $0.00 | Parse and Extract, Team and Enterprise plans |
| Ground | no credits | $0.00 | Still logged in usage history |
Extract reads markdown, so the parse charge comes first and Extract is added on top.
Two rules in the Gen2 credit docs decide your bill more than the model you pick. How you call the API, and how much text your pages return.
A synchronous Parse call runs on the priority tier every time. An async Parse Job runs on the standard tier unless you ask for priority, and standard is half the rate. So one typical page on DPT-3 Pro costs 3 credits through the sync endpoint and 1.5 credits as a job. Teams that wire up the sync endpoint for batch work pay double for speed they never use.
Gen1 charged a flat 3 credits per page. Gen2 charges a per-page rate plus a rate per 1,000 characters of returned markdown, counting page-break and document-ID markers. On DPT-3 Pro priority, the output charge on a typical page (2 credits) is already twice the page charge (1 credit). The text on the page, not the page count alone, drives the bill.
A light form that returns little text stays near the per-page rate. A typical 4,000-character page on DPT-3 Pro priority is $30 per 1,000 pages. A dense 8,000-character page, such as a statement full of tables, is $50 per 1,000. Budget on your own documents, not on a single per-page number.
Landing AI documents a 12-page synchronous DPT-3 Pro parse that returns 48,120 output characters. Here is how the bill adds up, next to what the same document costs as a standard job.
| Step | Sync call (priority) | Async job (standard) |
|---|---|---|
| Pages | 12 x 1 = 12 credits | 12 x 0.5 = 6 credits |
| Output characters | 48,120 / 1,000 x 0.5 = 24.06 | 48,120 / 1,000 x 0.25 = 12.03 |
| Total, rounded up | 36.06 to 36.1 credits | 18.03 to 18.1 credits |
| Cost | $0.361 | $0.181 |
The sync column is the example as Landing AI publishes it. The job column applies the standard-tier rates to the same document. The response metadata reports page count, output characters, service tier and total credits, so you can check every charge.
A credit costs $0.01 on Explore and Team, and the price is the same for US and EU accounts. Plans change who can use the account and what compliance you get, not the per-page rates above.
Pay as you go
Starts with 1,000 free credits that expire 90 days after you create the account. One user. After the free credits, Explore runs only on credits you buy, and prepaid credits expire one year after purchase.
From $250 a month
Four sizes from 25,000 to 200,000 credits a month, unlimited users, zero data retention and a HIPAA BAA. At $250, 25,000 credits buys about 16,600 typical pages as DPT-3 Pro standard jobs, or about 8,300 through sync calls.
Custom
For more than 200,000 credits a month, with VPC and on-prem deployment, an SLA and SSO. Pricing is set by contract.
Gen1 is the pre-DPT-3 generation, and Landing AI still publishes its rates. Parse is a flat charge per page with no output-character component. The DPT-1 and DPT-2 mini models now sit under Legacy ADE Features and are deprecated.
| Gen1 operation | Credits | Per 1,000 pages | Notes |
|---|---|---|---|
| Parse (documents) | 3 credits / page | $30.00 | Flat per page, no output-character charge |
| Zero data retention on Parse | +1 credit / page | +$10.00 | Team and Enterprise plans |
| Parse (spreadsheets) | 1 credit / sheet + 3 / embedded image | by sheet | Images, charts and logos count as embedded images |
| Extract | 1 credit / 5,000 input chars + 1 / 1,000 output chars | by character | Same rate sync or async |
| Classify | 0.5 credits / page | $5.00 | Document type only, not parsing |
| Split | 1 credit / 5,000 input chars | by character | Finds document boundaries |
| Section | 1 credit / 5,000 input chars + 1 / 1,000 output chars | by character | Builds a table of contents |
The 0.5 credits per page that often gets quoted for Landing AI is the Gen1 Classify rate, which only labels the document type. It is not a parse price. Gen1 Classify and the character-billed Gen1 operations round up to the nearest tenth of a credit.
Landing AI wins on one specific job. For digitally created Latin-script documents sent as standard jobs, DPT-3 Verity at $5.50 per 1,000 typical pages is cheaper than Reducto r-1 Parse at $10 and LlamaParse Agentic at $12.50. DPT-3 Pro, the default model, costs more than both, and LlamaParse Fast and Cost-effective cost less per page than any Landing AI tier. Reducto r-1 only bills $10 when the request names the model; otherwise legacy Parse rates apply. 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 | Notes | Free usage or entry plan |
|---|---|---|---|
| Landing AI DPT-3 Verity, standard job | $5.50 | Typical page, digital Latin-script documents, preview model | 1,000 credits, expire in 90 days |
| Landing AI DPT-3 Pro, standard job | $15.00 | Typical page, default model | 1,000 credits, expire in 90 days |
| Landing AI DPT-3 Pro, sync call | $30.00 | Typical page, $50 for a dense 8,000-character page | 1,000 credits, expire in 90 days |
| Landing AI Gen1 Parse | $30.00 | Older generation, flat per page | 1,000 credits, expire in 90 days |
| LlamaParse Fast | $1.25 | Lowest parse tier, 1 credit per page | 10,000 credits a month (Free plan) |
| LlamaParse Cost-effective | $3.75 | 3 credits per page | 10,000 credits a month (Free plan) |
| LlamaParse Agentic | $12.50 | 10 credits per page | 10,000 credits a month (Free plan) |
| LlamaParse Agentic Plus | $56.25 | 45 credits per page | 10,000 credits a month (Free plan) |
| Reducto r-1 Parse | $10.00 | Preview, V3 API, only when the request sets the model to r-1 | $150 (15,000 credits), one time |
| Reducto legacy Parse | $15 to $60 | Standard or complex page, doubled in agentic modes | $150 (15,000 credits), one time |
| Reducto Extract | $20.00 | Parsing included | $150 (15,000 credits), one time |
| Cloud OCR (Textract, Azure, Google) | $1.50 | Plain text OCR | Varies by provider |
| DocuOCR | $14.26 to $19.60 | Classification, extraction, validation, review and export included | Starter $49 a month for 2,500 pages |
Landing AI rows use a typical 4,000-character page. LlamaParse and Reducto r-1 quote a flat rate per page. DocuOCR is the monthly plan price divided by included pages.
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
