INSERT

The INSERT statement adds new rows to a table.

Warning

INSERT is experimental and only works against local or limited storage backends. It is not suitable for production use.

Syntax

INSERT INTO <table_name> [ ( <column> [, ...] ) ]
{ VALUES ( <value> [, ...] ) [, ...] | <select_statement> };

Parameters

<table_name> — the table to insert into.
<column> — an explicit column list. Optional; when given, it must name every column in the target table — see Partial Inserts below. Omitting it inserts into the table's columns in schema order.
<value> — literal values for one row. Give more than one parenthesized group, comma-separated, to insert multiple rows in one statement.
<select_statement> — a SELECT query whose result rows are inserted in place of a VALUES list.

Examples

Single Row Insert

INSERT INTO users (id, name, email, active)
VALUES (1, 'John Doe', 'john@example.com', TRUE);

Multiple Row Insert

INSERT INTO users (id, name, email, active)
VALUES 
  (1, 'John Doe', 'john@example.com', TRUE),
  (2, 'Jane Smith', 'jane@example.com', TRUE),
  (3, 'Bob Johnson', 'bob@example.com', FALSE);

Insert Without a Column List

Values are matched to the table's columns in schema order:

INSERT INTO users
VALUES (1, 'John Doe', 'john@example.com', TRUE);

Insert from SELECT

INSERT INTO users_backup (id, name, email)
SELECT id, name, email FROM users WHERE archived = FALSE;

Partial Inserts

Inserting into a subset of a table's columns is not supported. An explicit column list must name every column in the target table; a shorter list is rejected when the query is planned, rather than leaving the unlisted columns to be filled with something you did not ask for:

INSERT explicit column list must list all target columns
(target has 4, got 3). Partial column inserts are not yet supported.

The list may reorder columns relative to the schema — it just cannot omit any.

Notes

INSERT is experimental and only works against local or limited storage backends.
The column list is optional; when present it must be complete (see above).
Column order in the VALUES clause must match the column list, or the table's schema order when no list is given.
The number of values per row, and each value's type, are checked at plan time against the target's schema — a mismatch is rejected before any data is written.
INSERT OVERWRITE is not supported.
A materialized view is not a table: this statement is rejected against one. Its contents come from its defining SELECT — see REFRESH MATERIALIZED VIEW.