Text to table
extractor
Paste any messy text - an email thread, a scribbled list, an invoice dump. It finds the emails, phones, links, dates, amounts and names, highlights every one, and lines them up in a clean table you can export.
Watch it become a table
Yetty turns lists like this one into website collections - a team page, a price list, a directory - editable by you or your AI.
What it detects,
and how
Emails and links first
The most unambiguous patterns are claimed first, so a phone regex never eats half of an email address and a URL keeps its slashes and query string intact.
Phones, international
Country codes with +, parentheses area codes, dots, dashes or spaces - anything with 7 to 15 digits in a phone-like arrangement is caught after dates and amounts are already out of the way.
Dates become ISO
3/2, Mar 3 2026, 15.4.26 and 2026-04-15 all normalize to yyyy-mm-dd. A missing year assumes the current one; a day above 12 flips the day/month reading automatically.
Money keeps its currency
Symbols and codes both count: $40, 40 USD, EUR 12.50, 99 ILS. The table stores the amount with its currency code so a mixed-currency list stays honest.
Names are a guess
Capitalized word pairs that survive after everything else is extracted are offered as names. It is the softest heuristic of the set - which is exactly why the highlight view lets you verify it.
Records split themselves
Blank-line paragraphs, bullet lists and plain line-per-record lists are recognised automatically, and the chip above the table tells you which reading it chose.
People also ask
How do I turn a messy list of names and emails into a spreadsheet?
Paste the list here - one contact per line, bullets, or paragraphs, it does not matter. The extractor finds the emails, phone numbers, dates, links and amounts in each record, highlights them so you can check its work, and lines everything up in a table you can export as CSV for Excel or Google Sheets. It all happens in your browser, so a client list never leaves your machine.
Can it extract phone numbers and dates written in different formats?
Yes. Phones are matched loosely enough to catch +972 54 123 4567, (202) 555-0148 and 054-1234567 alike. Dates are recognised in many shapes - 2026-04-15, 3/2, Mar 3 2026, 15.4.26 - and every one is normalised to the same ISO yyyy-mm-dd form in the table, so the column actually sorts correctly once you open it in a spreadsheet.
Is this the same as asking an AI to extract a table from text?
It is the deterministic cousin. Everything here is pattern-matching that runs locally: fast, free, private, and it never invents values that were not in your text. An AI is better at understanding meaning ("which of these people already paid?"), but for pulling out the emails, phones, dates and amounts that are literally written in the text, local extraction is instant and trustworthy.