Send a lease PDF or scan to a REST endpoint and get landlord, tenant, dates, base rent and the rent schedule back as JSON, each value tied to its source page.
Upload a document to extract
Drop files here or click to upload
Up to 50 files at once, 50MB each, zip archives up to 190MB
Uploading...
Each route ends with structured lease data. They differ in who reads the lease and how much engineering you carry.
| Route | What it requires | Effort or cost | Where it falls short |
|---|---|---|---|
| Keep abstracting by hand | Analysts and a shared drive | 4 to 8 analyst hours per complex lease | Does not scale to a portfolio onboarding, and nothing is machine readable until someone types it |
| Offshore abstraction service | A vendor contract and email hand-offs | Priced per lease, turnaround in days | No API for your product, and every amendment is another order |
| Enterprise lease platform with an API | A platform license and an implementation | Quote based, long onboarding | The API usually reads data already inside the platform, not a raw lease PDF |
| Build on a general LLM yourself | Prompts, OCR, queues, evaluation | Engineering months plus ongoing tuning | You own page citations, retries, scans and every regression |
| LeaseAbstractors API This tool | An API key on a Pro plan | Three calls, minutes per lease | You map the JSON to your system and still review key terms before posting |
Effort figures are typical analyst time for a commercial lease with amendments. API limits are taken from the LeaseAbstractors API documentation.
The pieces a developer checks before committing to an extraction vendor.
Upload a lease, start extraction, then fetch the result or receive it by webhook. Three calls take a PDF to structured JSON.
The result includes a field_pages map from each abstract field to the 1-based page it was read from, so your reviewers can verify a value without opening the whole lease.
Extraction returns an extraction_hash immediately. Register an HTTPS endpoint and the finished payload is posted to it, signed with an HMAC-SHA256 header you verify before processing.
PDF, JPG, PNG, BMP, HEIC and TIFF up to 50 MB, including multi-page scans and amendments.
List built-in and team templates, then pass a document_type_id to extract your own field schema, or leave it out and let the AI detect the document.
Pro allows 500 requests per hour per key and 10 concurrent jobs, with standard rate-limit headers and a Retry-After on 429. Enterprise adds far more monthly page volume.
A working integration usually takes an afternoon, most of it spent mapping fields.
Upgrade to Pro, open Settings then API Keys and create a key. It is shown once, so store it in your secrets manager.
POST the PDF or scan to the upload endpoint and keep the file_id it returns.
POST the file_id, optionally with a document_type_id. You get an extraction_hash back with a 202 response.
Poll the results endpoint or let the webhook deliver structured JSON with values, rent schedule rows and source pages.
Who it is for, how a job runs and what it leaves to you.
Short answer: LeaseAbstractors exposes a REST API on its Pro and Enterprise plans. You send a lease PDF or scan, start an asynchronous extraction job and receive structured JSON with landlord, tenant, dates, base rent, rent schedule rows and the page each value came from. It is built for teams that want abstraction inside their own product or pipeline, not another screen to log into.
Three buyers come up most. PropTech and lease software companies that need to turn customer lease PDFs into records without building OCR and extraction in house. Accounting and lease accounting teams that want abstracted terms to land in a ledger or ASC 842 model automatically. And asset managers or lenders running diligence on hundreds of leases, who want every file processed the same way with a page citation that a reviewer can check.
Upload sends the file as multipart form data and returns a file_id. Extract takes that file_id and an optional document_type_id and answers with HTTP 202 and an extraction_hash. Results returns the finished job. While it runs, the status reads pending or processing, and when done it reads completed or failed. Every request carries a bearer token in the Authorization header, and every error response includes a machine readable error_code.
The result is shaped for software. Plain fields come back as content and value pairs with a page number. Tables such as the rent schedule come back as rows of cells, so you can write them straight into a payment schedule. A field the model could not place on a page is left out of the field_pages map, which means a missing key is honest, never a default of page 1.
Polling works for a few documents. For a portfolio onboarding, register an HTTPS endpoint and let the extraction.completed event arrive on its own. The delivery is signed with an HMAC-SHA256 header, so your service can reject anything that did not come from us. The payload carries the extraction hash, the document type and the fields.
It extracts and cites. It does not push to Yardi, MRI, SAP or NetSuite on its own, so you map the JSON fields to your system and write them in. It also does not replace review. For leases that feed lease accounting, have a person check rent steps, options and notice dates against the cited pages before the data is posted.
API access sits on the Pro plan, with Enterprise for large monthly page volume. Read the API documentation first, it lists every endpoint with request and response examples in cURL, PHP, JavaScript and Python, then pick a plan on the pricing page. To judge output quality before you integrate, run a few of your own leases through the upload box above and compare the fields with the originals.
If you are still deciding between building and buying, read the comparison of lease abstraction software, or see how services compare with software for a one off project.
Still have questions? Our team is happy to help.
Talk to our teamYes. Pro and Enterprise plans include a REST API. You upload a lease, start an extraction job and retrieve structured JSON with key terms, rent schedule rows and the source page of each value. Full endpoint documentation is public.
A JSON object with a status, the document type, a list of fields with value and page, parsed tables such as the rent schedule, and a field_pages map. Typical fields are landlord, tenant, commencement date, expiration date and base rent.
API access is part of the Pro plan and above, including Enterprise. Free and lower tiers can use the web upload but cannot create API keys. Check the pricing page for current plan prices before you build against it.
No. Extraction is asynchronous. The extract call returns 202 Accepted with an extraction_hash, and the job moves through pending, processing and completed or failed. You can poll the results endpoint or register a webhook for the extraction.completed event.
Each delivery carries an X-LeaseAbstractors-Signature header, an HMAC-SHA256 of the raw request body using the signing secret from your dashboard. Verify it before you trust the payload, and serve the endpoint over HTTPS.
Pro keys are limited to 500 requests per hour, with a 50 MB file limit and 10 concurrent jobs. Responses include X-RateLimit headers, and a 429 includes Retry-After. Enterprise is built for higher monthly page volume.
Yes. List available document types, including custom templates your team configured, and pass the document_type_id when you start extraction. Omit it and the AI detects the document type itself.
Yes. Scans and photos are read with OCR before extraction, and PDF, JPG, PNG, BMP, HEIC and TIFF files are accepted. Always review low quality scans, because a faint stamp or handwritten rider can change a value.
Load a whole portfolio folder in one session.
Learn moreThe full list of fields the extraction reads.
Learn moreHow the model reads and cites a lease.
Learn moreFeed abstracted lease data to your ledger.
Learn moreNotice windows and option dates as data.
Learn moreCompare the tools before you integrate.
Learn moreFor individuals and small teams beginning to automate document data capture.
USD / mo
$288 charged today
With : due today
Choose extraction speed vs. model accuracy per job.
| Base AI Faster | 2,500 pages |
| Pro AI Best accuracy | 500 pages |
For growing teams that need bulk extraction, custom templates and spend analytics.
USD / mo
$888 charged today
With : due today
Choose extraction speed vs. model accuracy per job.
| Base AI Faster | 10,000 pages |
| Pro AI Best accuracy | 2,000 pages |
High-volume extraction with API access, priority processing, and adjustable page tiers.
USD / mo
$ charged today
With : due today
Scales with your selected tier.
| Base AI Faster | pages |
| Pro AI Best accuracy | pages |
For portfolio-scale teams abstracting hundreds of thousands of pages a month, with unlimited seats.
USD / mo
$41,880 charged today
With : due today
| Base AI Faster | 500,000 pages |
| Pro AI Best accuracy | 120,000 pages |
Everything you need to know before you subscribe, or ask our support team about anything specific.
Ask a questionEach plan includes a monthly allowance of pages, and the number you get depends on which AI model you choose at extraction time. Base AI processes pages faster; Pro AI delivers the highest field-level accuracy. Your allowance resets on your billing anniversary each month, unused pages do not roll over. A typical commercial lease runs 20 to 60 pages including amendments, so the Starter plan's 2,500 Base AI pages cover roughly 40 to 120 leases a month.
Once your included pages are used up, extraction pauses until the next billing cycle. You can upgrade your plan at any time mid-month, the higher allowance kicks in immediately and you are only charged the prorated difference for the remainder of the period.
Yes. You can switch from monthly to yearly billing at any time on the Billing page. The yearly price is charged the same day, minus the unused part of your current month, and the 50% yearly rate applies right away. All payments are final: cancelling stops the next renewal, and access continues until the end of the period already paid.
All plans export to Excel XLSX and CSV. Plus adds QuickBooks (QBO) export and financial reports. Pro unlocks JSON output via REST API so you can push structured data directly into any ERP, RPA platform or internal system.
We do not use your documents to train or improve our own models, and files are encrypted in transit (TLS 1.2+) and at rest (AES-256). Extraction runs through a third-party model provider, so review their terms if your policy requires a contractual no-training commitment. You control retention: documents can be deleted from the workspace at any time, and we will remove everything on request.
Enterprise costs $6,990 per month billed monthly, or $3,490 per month billed yearly ($41,880 charged today). It includes 500,000 Base AI pages and 120,000 Pro AI pages a month, 1,000 AI Smart Reports, unlimited user seats and everything in Pro, including REST API access. You buy it online like any other plan, and you can move between Pro and Enterprise at any time from your billing page, paying only the prorated difference.
Enterprise gives you 500,000 Base AI pages and 120,000 Pro AI pages a month, unlimited seats and API access, from $3,490 per month billed yearly. Buy it online and start right away.
