Rent Roll API

One endpoint does the work: upload a rent roll, get structured data back with a deterministic verification report. No async jobs, no polling — a single HTTP call.

Machine-readable spec:OpenAPI 3.0 (openapi.json)— import into Postman, Insomnia, or your codegen tool.

Pass your API key in theX-Api-Keyheader (or as a bearer token). Get a key from thepricing page; manage it atyour account.

X-Api-Key: rr_live_your_key_here # or Authorization: Bearer rr_live_your_key_here

curl -X POST "https://rentrollapi.com/v1/extract?format=json" \ -H "X-Api-Key: rr_live_your_key_here" \ -F "file=@rent-roll.pdf"

Official client on NuGet:RentRollAPI.Client(netstandard2.0 / net8.0).

using RentRollApi.Client; using var client = new RentRollClient("rr_live_your_key_here"); var result = await client.ExtractAsync("rent-roll.pdf"); Console.WriteLine($"{result.Units.Count} units, confidence {result.Verification.DocumentConfidence:P0}"); foreach (var check in result.Verification.Checks) Console.WriteLine($"[{check.Result}] {check.Name}: {check.Detail}"); // Or get Excel / CSV directly: byte[] xlsx = await client.ExtractToXlsxAsync(stream, "rent-roll.pdf");

Every field carries a value, status (Extracted/Missing/Flagged), and confidence. The verification block is computed deterministically — totals are re-added and reconciled against what the document states, so you know when to trust the output.

{ "propertyName": { "value": "Maplewood Commerce Center", "status": "Extracted", "confidence": 0.9 }, "asOfDate": { "value": "2026-06-30", "status": "Extracted", "confidence": 0.9 }, "units": [ { "unitNumber": { "value": "101", "status": "Extracted", "confidence": 0.9 }, "tenantName": { "value": "Cedar Point Dental", "status": "Extracted", "confidence": 0.9 }, "squareFeet": { "value": 1850, "status": "Extracted", "confidence": 0.9 }, "monthlyRent": { "value": 4625.0, "status": "Extracted", "confidence": 0.9 }, "leaseStart": { "status": "Missing" }, "leaseEnd": { "value": "2028-03-31", "status": "Extracted", "confidence": 0.9 }, "occupancy": { "value": "Occupied", "status": "Extracted", "confidence": 0.9 } } ], "totals": { "unitCount": 8, "monthlyRent": 28155.0, "squareFeet": 13750 }, "verification": { "documentConfidence": 0.9, "flaggedFieldCount": 0, "allChecksPassed": true, "checks": [ { "name": "UnitCountReconciliation", "result": "Pass", "detail": "Extracted 8 units; document states 8." }, { "name": "TotalMonthlyRentReconciliation", "result": "Pass", "detail": "Computed 28,155.00 vs stated 28,155.00 (within tolerance)." }, { "name": "TotalSquareFeetReconciliation", "result": "Pass", "detail": "Computed 13,750.00 vs stated 13,750.00 (within tolerance)." }, { "name": "DuplicateUnits", "result": "Pass", "detail": "No duplicate unit numbers." }, { "name": "LeaseDateOrder", "result": "Pass", "detail": "All lease end dates follow start dates." }, { "name": "NonNegativeRents", "result": "Pass", "detail": "No negative rents." } ] } }

All errors return{"error": "human-readable message"}.

Extract a T12 / operating statement (trailing-twelve-month income & expense report) into structured line items. Same inputs, auth, metering, andformatoptions as/v1/extract— one extraction counts as one document against your quota.

curl -X POST "https://rentrollapi.com/v1/extract/operating-statement" \ -H "X-Api-Key: rr_live_your_key_here" \ -F "file=@t12.pdf"

{ "propertyName": { "value": "Maplewood Apartments", "status": "extracted", "confidence": 0.95 }, "basis": { "value": "Accrual", "status": "extracted", "confidence": 0.9 }, "months": ["2025-08", "2025-09", "...", "2026-07"], "lineItems": [ { "label": { "value": "Gross Potential Rent", "status": "extracted", "confidence": 0.95 }, "kind": "Income", "category": "GrossPotentialRent", "monthlyValues": [ { "value": 52000, "status": "extracted", "confidence": 0.9 }, "..." ], "total": { "value": 631200, "status": "extracted", "confidence": 0.95 } } ], "verification": { "documentConfidence": 0.9, "allChecksPassed": true, "checks": [ { "name": "RowTotalReconciliation", "result": "pass", "detail": "All 15 row totals reconcile with their monthly values." }, { "name": "NoiReconciliation", "result": "pass", "detail": "EGI − OpEx matches stated NOI (within tolerance)." } ] } }

Deterministic checks: row totals vs monthly values, EGI = income sum, Total OpEx = expense sum, NOI = EGI − OpEx, non-negative expenses, and consecutive-month continuity. Line-itemkindisIncome,Expense,Subtotal, orOther;categorynormalizes labels into ~23 standard buckets (RealEstateTaxes, Insurance, Utilities, RepairsMaintenance, Payroll, ManagementFee,...). Contra-revenue rows like vacancy and concessions are Income-kind with negative values.

RentRollAPI is also aModel Context Protocolserver. Point any MCP-capable client (Claude, ChatGPT, Copilot, Cursor...) at the streamable-HTTP endpoint and it gainsextract_rent_rollandextract_operating_statementtools:

{ "mcpServers": { "rentrollapi": { "type": "http", "url": "https://rentrollapi.com/mcp", "headers": { "X-Api-Key": "rr_live_your_key_here" } } } }

Documents are processed in memory and never stored after your extraction completes. See theprivacy policy.

Official MCP bridge to the Invoket x402 gateway: 68 pay-per-call endpoints (IBAN validation and sanctions screening, phone validation, French and EU law in force, weather and climate, medication data) exposed as typed tools, paid in USDC on Base with non-custodial BYOK spend caps.