A free, open-source desktop app for querying the data files on your computer with SQL. It runs DuckDB inside the app: drop in CSV, Parquet, JSON or Excel files and each becomes a table you can query, with no server, database setup or account.

Does DuckLocal upload my data?

No. Files are read where they are, on your computer, and are never copied, uploaded or changed. DuckLocal has no account, no telemetry and no cloud service.

Which file formats can it open?

CSV, TSV, Parquet, JSON and newline-delimited JSON, Excel (.xlsx, .xls, .xlsb) and OpenDocument (.ods) spreadsheets, DuckDB database files, whole folders, glob patterns such as *.csv, and objects in Amazon S3 or S3-compatible storage.

Which systems does it run on?

macOS 12 or later on Apple silicon, 64-bit Windows 10 or later, and x86_64 Linux with glibc 2.35 or newer. On an Intel Mac, build it from source.

Yes. It is free and open source under the Apache-2.0 license, with the full source code on GitHub.

How large can my data be?

Queries run in DuckDB directly over the files, so nothing is loaded into the app first. The results grid shows up to 100,000 rows or 2,000,000 cells of a result; exporting to CSV or Parquet re-runs the query and writes the full result.

Can AI agents like Claude Code or Codex query my files with it?

Yes. The ducklocal command runs one SQL statement from the terminal and prints the result as JSON, with no window and no server. The official agent skill teaches an agent to inspect the schema before it queries.

Can it query data in S3?

Yes. DuckLocal uses DuckDB's httpfs extension to browse buckets and query files in Amazon S3 and S3-compatible storage. Credentials are kept for the session only.

DuckLocal — SQL on the files on your computer

#	channel	revenue
1	Mobile app	290549.76
2	Web	289544.8
3	Store	240067.6
4	Phone	82511.2

From a file to an answer

No connection to set up and no schema to write first.

Drop in a file

Drag in a CSV, a spreadsheet or a whole folder. No import step, no schema to write.
- Local files1
- sales224
- dateDATE
- channelVARCHAR
- revenueDOUBLE
It’s a table already

sales.csv becomes the table sales, with its columns and types in the sidebar. The file itself is never copied or changed.
Ask with SQL

Press ⌘↵ for the answer, then chart it or save it as CSV or Parquet.

Opens what you already have

Everything opens where it is, and comes back on the next launch.

.csvComma-separated
.tsvTab-separated
.parquetColumnar, fast to read
.json .ndjsonDocuments, or one record per line
.xlsx .xls .odsOne table per sheet
.duckdbOpens as the database
s3://Buckets and S3-compatible storage
folder/Every data file inside, or a pattern like *.csv

Let your AI agent ask

Point an agent at your files and it can answer questions about them. ducklocal runs one SQL statement from the terminal and prints the result as JSON — no window, no server. The official agent skill teaches it to read the schema before it queries.

CLI and agent skill

Which channel brought in the most revenue?

Terminal

$ ducklocal query --sql "SELECT channel, round(sum(revenue), 2) AS revenue FROM 'sales.csv' GROUP BY channel ORDER BY revenue DESC"
{"columns":[{"name":"channel","type":"Utf8"},{"name":"revenue","type":"Float64"}],"rows":[["Mobile app",290549.76],["Web",289544.8],["Store",240067.6],["Phone",82511.2]],"row_count":4,"truncated":false,"elapsed_ms":14}

Mobile app, with 290,549.76 — just ahead of Web at 289,544.80. Phone is far behind at 82,511.20.

Dashboards as plain files

Describe a chart in a few lines of a .dash file, and DuckLocal draws it as a dashboard tab. Plain text means it diffs, reviews and validates: ducklocal check catches a missing column before anyone opens it.

Dashboards (.dash)

query "by_channel" {
  sql = <<SQL
    SELECT channel, sum(revenue) AS revenue
    FROM sales GROUP BY channel
    ORDER BY revenue DESC
  SQL
}

plot "revenue" {
  type  = "bar"
  query = query.by_channel
  x     = channel
  y     = revenue
  title = "Revenue by channel"
}

Your data stays yours

DuckLocal runs DuckDB inside the app, on your computer. There is nothing to sign up for and nowhere for your files to go.

Files are read where they are — never copied, never uploaded.
No account, no telemetry, no cloud service.
Dashboards can only read, so opening one someone sent you cannot change your data.
Open source under Apache-2.0: every line is there to read.

Questions people ask

What is DuckLocal?: A free, open-source desktop app for querying the data files on your computer with SQL. It runs DuckDB inside the app: drop in CSV, Parquet, JSON or Excel files and each becomes a table you can query, with no server, database setup or account.
Does DuckLocal upload my data?: No. Files are read where they are, on your computer, and are never copied, uploaded or changed. DuckLocal has no account, no telemetry and no cloud service.
Which file formats can it open?: CSV, TSV, Parquet, JSON and newline-delimited JSON, Excel (.xlsx, .xls, .xlsb) and OpenDocument (.ods) spreadsheets, DuckDB database files, whole folders, glob patterns such as *.csv, and objects in Amazon S3 or S3-compatible storage.
Which systems does it run on?: macOS 12 or later on Apple silicon, 64-bit Windows 10 or later, and x86_64 Linux with glibc 2.35 or newer. On an Intel Mac, build it from source.
Is DuckLocal free?: Yes. It is free and open source under the Apache-2.0 license, with the full source code on GitHub.
How large can my data be?: Queries run in DuckDB directly over the files, so nothing is loaded into the app first. The results grid shows up to 100,000 rows or 2,000,000 cells of a result; exporting to CSV or Parquet re-runs the query and writes the full result.
Can AI agents like Claude Code or Codex query my files with it?: Yes. The ducklocal command runs one SQL statement from the terminal and prints the result as JSON, with no window and no server. The official agent skill teaches an agent to inspect the schema before it queries.
Can it query data in S3?: Yes. DuckLocal uses DuckDB's httpfs extension to browse buckets and query files in Amazon S3 and S3-compatible storage. Credentials are kept for the session only.

Try it on your own files

Download DuckLocal, drop in a file and run a query. The 10-minute tutorial walks through it with a sample file.

Download for Mac 10-minute tutorial