LimAI Docs

This guide walks you through the complete workflow to extract structured data from a document using the LimAI API.

Prerequisites

A LimAI account with an API token
An extraction schema (model) created in the LimAI dashboard
A document to process (PDF, image, or Excel file)

Step 1: Get an Upload URL

Request a pre-signed URL and a fileId for your document.

const response = await fetch(
`https://app.limai.io/api/v1/document/${SCHEMA_ID}/get-url?filename=invoice.pdf`,
{
  headers: { "Authorization": `Bearer ${API_TOKEN}` }
}
)
const { url, fileId } = await response.json()

Step 2: Upload Your File

Upload the file directly to the pre-signed URL using a PUT request.

const fs = require("fs")
const fileData = fs.readFileSync("invoice.pdf")

await fetch(url, {
method: "PUT",
headers: { "Content-Type": "application/pdf" },
body: fileData
})

Step 3: Process the Document

Trigger extraction on the uploaded file. Use the synchronous endpoint for small files or the async endpoint for larger documents.

Synchronous (waits for result)

const result = await fetch(
`https://app.limai.io/api/v1/document/${SCHEMA_ID}/process-file/${fileId}`,
{
  method: "POST",
  headers: {
    "Authorization": `Bearer ${API_TOKEN}`,
    "Content-Type": "application/json"
  },
  body: JSON.stringify({})
}
)
const extracted = await result.json()

Asynchronous (returns job ID, poll for results)

const queueResponse = await fetch(
`https://app.limai.io/api/v1/document/${SCHEMA_ID}/process-file-async/${fileId}`,
{
  method: "POST",
  headers: {
    "Authorization": `Bearer ${API_TOKEN}`,
    "Content-Type": "application/json"
  },
  body: JSON.stringify({})
}
)
const { jobId } = await queueResponse.json()

Step 4: Retrieve Results

For async processing, poll the get-file-data endpoint until the status is COMPLETED.

async function pollForResults(fileId) {
while (true) {
  const response = await fetch(
    `https://app.limai.io/api/v1/document/get-file-data?fileId=${fileId}`,
    { headers: { "Authorization": `Bearer ${API_TOKEN}` } }
  )
  const data = await response.json()

  if (data.status === "COMPLETED") return data
  if (data.status === "FAILED") throw new Error(data.errorMessage)

  await new Promise(r => setTimeout(r, 5000))
}
}

const result = await pollForResults(fileId)
console.log(result.data.tables)

Response200

{
  "status": "COMPLETED",
  "message": "File data retrieved successfully.",
  "fileId": "file_abc123",
  "data": {
    "tables": {
      "table_1": {
        "id": "table_1",
        "name": "Invoices",
        "columns": [
          {
            "id": "col_1",
            "name": "Invoice Number",
            "type": "TEXT"
          }
        ],
        "rows": [
          {
            "id": "row_1",
            "cells": {
              "Invoice Number": {
                "value": "INV-001",
                "columnId": "col_1"
              }
            }
          }
        ]
      }
    }
  }
}

Next Steps

Document Processing API -- Full endpoint reference
Classification API -- Automatically route documents to the right model
Webhooks -- Get notified when processing completes

Quick Start