What we're building next

New sourcesMore places to find businesses

• Apple Maps scraper

  Planned

  Local business lists from Apple Maps, same table, same enrichment as Google Maps.

• Bing Maps scraper

  Planned

  Local business lists from Bing Maps, so you can cross-check what Google Maps missed.

Next versionBuilt, in testing

• Undo, redo and fill down

  In testing

  ⌘Z and ⇧⌘Z step back through edits and pastes, ⌘D fills a value down a range, and pasting over a computed cell asks first.

• Big tables stay fast

  In testing

  Filters, search and sort run in a local database with infinite scroll, so tens of thousands of rows stay responsive.

• Cost per cell and per table

  In testing

  See provider, model, time taken and estimated cost for each cell, with today's and total cost in the status bar. AI columns show tokens used.

• Agent built into the app

  In testing

  A docked Claude Code or Codex terminal next to your table. Tables the agent creates open right in the app.

• Local Claude Code or Codex as the AI engine

  In testing

  Run AI columns through the agent already on your machine, one cell at a time, instead of paying per call to an API.

• Your own integrations

  In testing

  Describe any REST API once and it becomes a column. You, or your agent, can build it, and keys stay bound to the host they belong to.

• Hold for approval

  In testing

  Sends wait for you. See the exact request per row, then Approve or Reject. A rerun never sends the same row twice.

• Code column

  In testing

  Write a few lines of JavaScript to compute a cell, call an API, or ask AI, and save it as an integration.

• Typed columns

  In testing

  Text, paragraph, URL, email, phone, number and date, each rendered and linked properly.

• Column statistics and dependency hints

  In testing

  Fill rate, unique count and top values per column, and hover a header to see which columns feed it.

• Crawl, Search engine and Places search columns

  In testing

  Crawl a site from each row's URL into new rows, or run a Google search or Places lookup per row.

• Sturdier browser and background engine

  In testing

  The scraping engine restarts itself if it dies, so a crashed run recovers without you restarting the app.

Your turnSuggest a feature

Tell us what you wish Manycrawl did. The order this list gets built in is mostly decided by who asks, and we read every one.