Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/duckdb/duckdb/llms.txt

Use this file to discover all available pages before exploring further.

The DuckDB CLI can be configured using command-line options when starting the CLI, or by using dot commands within an interactive session.

​
Command-Line Options

Options can be specified when starting the DuckDB CLI: 
duckdb [OPTIONS] FILENAME [SQL]

​
Output Mode Options

Set the output format for query results:
  • -ascii - Set output mode to ‘ascii’ (uses ASCII unit and record separators)
  • -box - Set output mode to ‘box’ (Unicode box-drawing characters)
  • -column - Set output mode to ‘column’ (columnar output)
  • -csv - Set output mode to ‘csv’ (comma-separated values)
  • -html - Set output mode to ‘html’ (HTML table format)
  • -json - Set output mode to ‘json’ (JSON array)
  • -jsonlines - Set output mode to ‘jsonlines’ (newline-delimited JSON)
  • -line - Set output mode to ‘line’ (one column per line)
  • -list - Set output mode to ‘list’ (values delimited by |)
  • -markdown - Set output mode to ‘markdown’ (Markdown table format)
  • -quote - Set output mode to ‘quote’ (SQL-style quoting)
  • -table - Set output mode to ‘table’ (ASCII table)
Example: 
duckdb mydata.db -csv -c "SELECT * FROM users"

​
Execution Options

  • -c COMMAND - Run the SQL command and exit
  • -cmd COMMAND - Run the SQL command before reading stdin
  • -s COMMAND - Run the SQL command and exit (same as -c)
  • -f FILENAME - Read and process SQL commands from file and exit
Example: 
# Execute a command and exit
duckdb mydata.db -c "SELECT COUNT(*) FROM users"

# Execute a script file
duckdb mydata.db -f setup.sql

​
Input/Output Options

  • -echo - Print commands before execution
  • -header - Turn headers on
  • -noheader - Turn headers off
  • -interactive - Force interactive I/O
  • -batch - Force batch I/O (non-interactive mode)
  • -no-stdin - Exit after processing options instead of reading stdin
Example: 
# Show commands as they execute
duckdb mydata.db -echo -c "SELECT * FROM users"

# Batch mode with headers off
duckdb mydata.db -batch -noheader -c "SELECT * FROM users"

​
Initialization Options

  • -init FILENAME - Read initialization commands from the specified file
By default, DuckDB reads ~/.duckdbrc on startup. Use -init to specify a different file: 
duckdb mydata.db -init custom_init.sql

​
Database Options

  • -readonly - Open the database in read-only mode
  • -safe - Enable safe mode (disables external access)
  • -unsigned - Allow loading of unsigned extensions
  • -unredacted - Allow printing of unredacted secrets
Example: 
# Open in read-only mode
duckdb mydata.db -readonly

# Enable safe mode
duckdb mydata.db -safe

​
Separator Options

  • -separator SEP - Set output column separator (default: |)
  • -newline SEP - Set output row separator (default: \n)
  • -nullvalue TEXT - Set text string for NULL values (default: NULL)
Example: 
# Use comma as separator
duckdb -list -separator "," -c "SELECT * FROM users"

# Custom NULL representation
duckdb -nullvalue "N/A" -c "SELECT * FROM users"

​
Other Options

  • -bail - Stop after hitting an error
  • -version - Show DuckDB version and exit
  • -h or -help - Show help message and exit
  • -ui - Launch web interface using the UI extension
Example: 
# Show version
duckdb -version

# Exit on first error
duckdb mydata.db -bail -f script.sql

​
Output Formats

The CLI supports multiple output formats, configurable with the -mode option or .mode command.

​
Default (duckbox)

Unicode box-drawing with extensive formatting features: 
┌───────┬─────────┬───────┐
│  id   │  name   │  age  │
│ int32 │ varchar │ int32 │
├───────┼─────────┼───────┤
│     1 │ Alice   │    30 │
│     2 │ Bob     │    25 │
└───────┴─────────┴───────┘

​
CSV

Comma-separated values: 
id,name,age
1,Alice,30
2,Bob,25

​
JSON

JSON array format: 
[
  {"id":1,"name":"Alice","age":30},
  {"id":2,"name":"Bob","age":25}
]

​
JSON Lines

Newline-delimited JSON: 
{"id":1,"name":"Alice","age":30}
{"id":2,"name":"Bob","age":25}

​
Markdown

Markdown table format: 
| id | name  | age |
|----|-------|-----|
| 1  | Alice | 30  |
| 2  | Bob   | 25  |

​
List

Pipe-delimited values: 
id|name|age
1|Alice|30
2|Bob|25

​
Line

One value per line: 
     id = 1
   name = Alice
    age = 30

     id = 2
   name = Bob
    age = 25

​
Storage Version

Set the database storage compatibility version: 
duckdb -storage-version v0.10.0 mydata.db
This is useful for maintaining compatibility with older DuckDB versions.

​
Configuration File

DuckDB automatically reads ~/.duckdbrc on startup. This file can contain:
  • Dot commands (e.g., .mode, .headers)
  • SQL statements
  • Configuration settings
Example ~/.duckdbrc: 
-- Set default output mode
.mode duckbox

-- Enable headers
.headers on

-- Set null value display
.nullvalue NULL

-- Load common extensions
LOAD httpfs;
LOAD parquet;

-- Set up commonly used settings
SET threads TO 4;
SET memory_limit = '4GB';

​
Environment Variables

The CLI respects these environment variables:
  • DUCKDB_PAGER - Pager command to use (overrides PAGER)
  • PAGER - Default pager command (e.g., less, more)
  • EDITOR or VISUAL - Text editor for .edit command
  • DUCKDB_EDITOR - DuckDB-specific editor (overrides EDITOR and VISUAL)
Example: 
export DUCKDB_PAGER="less -SRX"
export DUCKDB_EDITOR="vim"
duckdb mydata.db

Build docs developers (and LLMs) love

Get started for freeTalk to us