A workbench for your data

Rows in.Answers out.

Open CSV, JSON and Parquet right in your browser. Filter, query and share the rows that matter.

No sign-upNo upload to explore

.csv.jsonl.json.parquet
Drop a file here
Stays in your browserUp to 1.5 GB

01 Explore

The whole file.
Just the rows you need.

Open the explorer
.csv

orders.csv

country:NL amount:>500
#order_idstatusamount
110482Refunded$842.00
210476Refunded$729.50
310471Refunded$685.00
410459Refunded$624.80
510443Refunded$589.00
1-5 of 5 rows

Filter without the fuss

Click a value, type a filter, or write SQL. Get straight to the subset you need.

See the shape of it

A distribution in every column. Spot the gaps, patterns and outliers at a glance.

Keep what matters

Export your view as CSV, JSON, Parquet or Markdown. Or share it with a link.

02 Keep the context

New rows.
Same place.

Push an updated file to the same link. See what's new, gone and changed. Pin a finding so the next person knows where to look.

The difference, made visible.

orders.csvv3 → v4

+312

new

-18

gone

47

changed

+refund_reason column

Send the finding, not the haystack.

Pinned finding38 rows

Refunds over $500 in NL

All from one merchant, since Sep 12.

country:NL amount:>500

03 For agents, too

From a command
to a conversation.

Your agent has the data. Give it a place to put it. One curl command or MCP call turns its output into something you can actually explore.

~/your-project
$ curl -T data.csv https://rows.page

https://rows.page/k3j9x2m1q7
48,210 rows, ready to explore.
claude mcp add --transport http rows https://rows.page/mcp

# Then ask your agent:
# Push results.csv and pin the failed rows as a finding.

No account needed. Anonymous links last 7 days.

Free to explore.
Pay to keep more.

Opening files is free forever. Plans are for what you keep hosted.

Compare plans