Dataset Summary Generator
Document a dataset in minutes. The generator analyses CSV or JSON and writes a Markdown summary – a “data card” – with the number of rows and columns, completeness, duplicates, the likely identifier column, the time span of date columns, a table describing every column, the main categories with their shares, and data quality notes. Add a name and source and paste it into a README, wiki or report.
- Runs in your browser
- No sign-up
- Free to use
Processed entirely in your browser – your data is not uploaded. Files up to 50 MB.
How to use Dataset Summary Generator
- Paste data or open a file.
- Enter the dataset name and source.
- Click Summarize dataset.
- Copy the Markdown into your documentation.
Dataset Summary Generator features
Markdown data card
README-ready.
Column table
Type, completeness, distinct, summary.
Categories
Shares of the main values.
Quality notes
Gaps, mixed types, outliers, duplicates.
Identifier and time span
Detected automatically.
Private
Runs in your browser; nothing is uploaded.
When to use Dataset Summary Generator
- Documenting exports for colleagues.
- Open data publishing.
- Handing over analysis projects.
- Data catalog entries.
Dataset Summary Generator FAQ
Does it describe what the data means?
No – it describes what is in the data. Add meaning, licence and collection method yourself.
Can I edit the result?
Yes, it is plain Markdown.
Which identifier is detected?
A unique, complete column whose name looks like an id, key, code or number.
Documentation people read
A short, factual summary at the top of a dataset saves every new user from profiling it again.
Private by design: your data is read and processed entirely in your browser with JavaScript. Nothing is uploaded to our server, so you can work with customer lists, exports and internal reports safely. Files up to 50 MB are supported; very large files depend on your device’s memory.
Input formats: paste or open CSV (comma, semicolon, tab or pipe – detected automatically), TSV or JSON (an array of objects, an object with a data, items or results array, or JSON Lines). The first CSV row is used as the header; JSON keys become columns, and nested values are kept as JSON text.
Missing values are recognised consistently across the data tools: empty cells and common placeholders such as NULL, N/A, NA, NaN, none, “-” and undefined. Types are inferred from the values – integer, decimal, boolean, date, email, URL, JSON or text – and columns with mixed types are flagged.
Related tools on this site cover the rest of a data workflow: the CSV cleaner, viewer, splitter and merger, converters between CSV, JSON, Excel and SQL, the JSON formatter and validator, and the other profiling, validation and statistics tools in this group.
Who it is for: analysts preparing data for a report, developers checking an API export or a database dump, marketers cleaning contact lists before an import, and students learning data quality. No account, no installation and no spreadsheet formulas are needed.