Export a dataset: CSV, JSON, JSON Lines, Markdown or Excel
Take a dataset’s rows out of Monity as CSV, JSON, JSON Lines, Markdown or Excel. The export has every row of the newest run, with the input each row came from.
Datasets are in early access. If your account doesn’t have them yet, open Datasets in the app and click Ask for early access, or write to us through the contact page.
Export the rows
- Open the dataset from Datasets.
- Click Export at the top of the page.
- Choose CSV, Excel, Markdown, JSON or JSON Lines. The file downloads, named after the dataset.
Export is there once the dataset has rows. It reads the newest run that collected rows, the same rows the Rows tab shows. To take the rows of an older run, open it on the Runs tab and click Rows as CSV.
The formats
| Format | What is in the file | Good for |
|---|---|---|
| CSV | A header row, then one line a row: Source (the input), Source address (its address), then the dataset’s fields. |
Spreadsheets, databases, BI tools |
| Excel | One sheet of rows. Numbers are numbers; everything else is text. | Sharing with people who work in Excel |
| Markdown | A heading with the dataset’s name, when it was collected, and a table of the rows. | Notes, tickets, and pasting into an AI assistant |
| JSON | An array with one object a row: source, sourceUrl and its values. |
Code and scripts |
| JSON Lines | The same objects, one a line. | Large datasets and streaming tools |
The field names are the ones you gave in the steps, and the values are as the transforms left them, such as a date as 2026-10-20. source names the input: its values for an input with columns, otherwise an id Monity gives the address. For a dataset of the Monity blog’s articles, a row in the JSON export looks like this:
{
"source": "in-2d0cf34bbf",
"sourceUrl": "https://monity.ai/blog",
"values": {
"url": "https://monity.ai/blog/best-media-pr-monitoring-tools/",
"date": "2026-10-20",
"title": "The best media and PR monitoring tools in 2026, and what each one actually watches",
"author": "Kamil",
"category": "Comparisons",
"read time": "11 min"
}
}A value copied from a web page can start with = and run as a formula when a spreadsheet opens it. CSV values are written so they open as plain text, and Excel cells are always text or numbers, never formulas.
Rows without exporting
- Every run, on its own: a Google Sheets or webhook alert sends the new, changed and gone rows each time a run finds some.
- Into another dataset: a dataset can start from this one’s rows, such as each article’s page (see Inputs and Find pages).
- In an AI assistant: Claude Code, Codex and other MCP clients read the rows as data or as a Markdown table. See Read datasets from Claude and Codex.