Free, in every format

Free

Finds tables and lists on its own

Press Scrape and Rowspell looks at the page you are on: tables, repeating lists such as product cards or search results, and article text. It ranks what it finds, marks the best match, and shows you a preview before it extracts anything.

Nothing is read until you press it. Rowspell does not scan pages on load.

Free

Point and pick when detection misses

If the list you want is not among the matches, pick it on the page: click two examples of the thing you want, then click the values that should become columns. Rowspell generalises from your examples to every item like them.

Free

Every page of a listing

"All pages" follows a listing's own next-page control, numbered pager or "load more" button, and infinite-scroll feeds are collected as you would scroll them. Duplicates across pages are removed.

A free run reads up to 50 pages, and only a page that produced rows counts toward that: a page that failed to load or was refused costs nothing. The limit is stated before the run starts, and a run that stops says where and why.

Free

Pages you are signed in to

Rowspell reads pages inside your own browser session, so a page you are logged into works like any other. It never reads your credentials. Banking, health and government pages ask for a typed confirmation before they are read.

Free

Clean it up before you export

Rows land in a grid where you can rename, retype, filter, sort and hide columns, edit cells and delete rows, with full undo. Columns are typed as text, number, currency, URL, date or rating, and a one-click cleanup tidies them.

Hover a row and the item it came from is outlined on the page. A long result opens in a full browser tab, and edits stay in sync between the tab and the panel. What you export is what you see.

Free

Export to CSV, Excel, JSON or NDJSON

Every format is free: CSV, XLSX, JSON, NDJSON, or copy straight into Google Sheets. CSV is written as UTF-8 with a guard against spreadsheet formulas; XLSX keeps numbers as numbers.

A large result can be split into 25,000-row files, each one complete on its own, with a manifest written last so you can tell a finished set from an interrupted one. Every row a run collected exports, whatever stopped the run.

Free

Recipes: save a scrape, run it again

Save what you picked as a recipe and re-run it later on the same site; the panel shows when a saved recipe matches the page you are on. Recipes are unlimited, and you can export them to a file and import them on another computer.

Free

Run one recipe over a list of URLs

Paste a list of URLs and Rowspell reads them one after another with the same fields, adding a URL column so every row says where it came from. A line it cannot use is shown with the reason, never dropped. Free runs read up to 50 URLs; every URL ends as read, empty, failed or refused, and the done screen shows which.

Free

Run history, and picking up where a run stopped

Every run leaves a record on your computer: what was read, when, and how many rows. The last 100 runs are kept. A list that stopped part way can be continued from where it stopped, as a new run.

Free

AI help with your own key

Optionally, connect an AI provider with your own key to generate a schema and name columns. Snippets are sent to that provider only when you ask, and the panel says so before you switch it on. Banking, health and government pages are refused outright.

Pro: more than one job

Pro

Detail crawl: the list, then each linked page

Choose the column that holds links, pick the fields you want on one real detail page, and Rowspell visits each link and adds those fields to the row it came from. Values are joined by URL, never by position, so a detail value can only land on its own row.

Pro

Scheduled runs and change alerts

Put a saved recipe on a daily or weekly schedule at a local time. Each run is compared with the last good one by a key column you choose, and you get one notification only when something was added, removed or changed, or when the run failed. Schedules run while Chrome is open; a run that was missed is recorded, not quietly run later.

Free and Pro, side by side

Free and Pro compared
FeatureFreePro
Pages per run50No limit
Rows per run10,000, stated before the run startsNo ceiling
FormatsCSV, XLSX, JSON, NDJSONThe same
Rows already collectedEvery one exports, in every formatThe same
RecipesUnlimitedUnlimited
Batch URL listsUp to 50 pages a runNo page limit
Scheduled re-runs with change alerts—Daily or weekly, while the browser is open
Detail crawl (list, then each detail page)—Included
Agent bridge (MCP)—Included
UpdatesAll 1.xAll 1.x

Get the data off the page.

Free in every format. No account. No telemetry.

See pricing