Enrich any CSV with AI and web search

The generic case. If you can describe the column you want in a sentence, and the answer exists on the public web, this works.

Updated August 2026

What goes in, what comes out

Input

  • Any identifying column

Output

  • Whatever you describe

The prompt

Paste this into the description field, or write your own version of it.

Describe the fields you want in plain English. OpenClay converts your description into a per-row prompt with your columns substituted in, and asks the model to return strict JSON.

Why this works

Nothing about the tool is specific to companies or sales. Every row is an independent prompt with that row's values substituted in, so the same machinery works for products, countries, papers, properties or anything else with a public footprint.

Getting the best results

Anything with a public footprint is fair game. The clearer your column names, the better the results: specific questions produce specific answers, and the five-row test shows you the quality before you commit to a batch.

What it costs

On Gemini 3 Flash, a typical row of this shape costs about <$0.01, so roughly $0.55 per thousand rows, paid directly to Google. OpenClay adds nothing.

The tool shows an exact figure after a five-row test, before you commit to the batch. See all models and pricing.

Step by step

  1. 1Upload the file. CSV, XLS or XLSX up to 10 MB, parsed entirely in your browser.
  2. 2Choose input columns. Pick the columns that identify the thing being researched.
  3. 3Name your output columns. Each becomes a new column in the downloaded file.
  4. 4Test, run, download. Five rows first for an exact cost, then the full batch.

No account, no card, no platform fee.

Bring your own API key and pay the model provider directly.

Start enriching, free

Other use cases