For the complete documentation index, see llms.txt. This page is also available as Markdown.

๐Ÿ“ŠCreate a Data Table

Generate spreadsheet-ready data tables by calling /content/Create with outputType: "datatable". Data table jobs are asynchronous: submit the job, poll /content/Status/{id}, then download the completed spreadsheet from /data-tables/{id}/download.

Step 1: Create the Data Table Request

curl -X POST "https://api.autocontentapi.com/content/Create" \
  -H "accept: application/json" \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "outputType": "datatable",
    "resources": [
      {
        "type": "website",
        "content": "https://example.com/market-report"
      }
    ],
    "text": "Extract the named companies, market segment, funding stage, region, and one-sentence positioning into a clean comparison table.",
    "language": "English"
  }'

Example response:

{
  "request_id": "550e8400-e29b-41d4-a716-446655440000",
  "status": 0
}

Save the request_id. It is also the data table ID used by the /data-tables endpoints.

Step 2: Poll for Status

While processing, treat any status from 1 to 99 as still running.

Step 3: Read the Completed Status Payload

When status reaches 100, the status response exposes the generated file URL in both datatable_url and share_url.

Use datatable_url when you only need the hosted file URL. Use the authenticated download endpoint below when you want the API to verify ownership and stream the file with download headers.

Step 4: Use the Data Tables API

List Data Tables

Example response:

The list includes in-progress data table requests. In-progress rows have status below 100 and url: null.

Download a Data Table

The API streams the generated spreadsheet as an attachment. The file extension is inferred from the stored asset and content type, usually .xlsx.

Supported Source Types

Data tables use the same source model as the general content endpoint:

Type
Description
Example

website

Public web page URL

"https://example.com/report"

youtube

YouTube video URL

"https://www.youtube.com/watch?v=VIDEO_ID"

text

Direct text source material

"Company A raised $10M..."

pdf

Base64-encoded PDF payload

"BASE64_ENCODED_PDF"

file

File ID returned by /files

"YOUR_FILE_ID"

You can also use topic, feedSelections, researches, podcastEpisodeIds, and projects when those sources are available in your account.

Prompting Tips

  1. Name the columns you want in text.

  2. Ask for normalized values when the source has inconsistent labels.

  3. Use a source with enough structured facts; data table generation works best when the input contains comparable entities, dates, metrics, categories, or rankings.

  4. Poll /content/Status/{id} before downloading. /data-tables/{id}/download returns 404 until the asset is ready.

Error Handling

If generation fails, /content/Status/{id} returns the same error metadata used by other asynchronous content jobs:

What's Next?

Last updated