TimescaleDB Column Types

TimescaleDB is built on PostgreSQL and supports all PostgreSQL column types. If a type is missing, open an issue (opens in a new tab).

Simple Types

These types are used without parameters:

Type	Aliases	Description
`bigint`	`int8`	8-byte signed integer
`bigserial`	`serial8`	Auto-incrementing 8-byte integer
`boolean`	`bool`	True/false value
`box`		Rectangular box on a plane
`bytea`		Binary data
`cidr`		IPv4 or IPv6 network address
`circle`		Circle on a plane
`citext`		Case-insensitive text
`date`		Calendar date
`double precision`	`float8`	8-byte floating point
`inet`		IPv4 or IPv6 host address
`integer`	`int`, `int4`	4-byte signed integer
`json`		JSON data
`jsonb`		Binary JSON data (indexable)
`line`		Infinite line on a plane
`lseg`		Line segment on a plane
`macaddr`		MAC address
`money`		Currency amount
`path`		Geometric path on a plane
`pg_lsn`		PostgreSQL Log Sequence Number
`point`		Geometric point on a plane
`polygon`		Closed geometric path on a plane
`real`	`float4`	4-byte floating point
`smallint`	`int2`	2-byte signed integer
`smallserial`	`serial2`	Auto-incrementing 2-byte integer
`serial`	`serial4`	Auto-incrementing 4-byte integer
`text`		Variable-length character string
`tsquery`		Text search query
`tsvector`		Text search document
`txid_snapshot`		Transaction ID snapshot
`uuid`		Universally unique identifier
`xml`		XML data

Parameterized Types

These types accept length, precision, or scale parameters:

Type	Example	Description
`bit(n)`	`bit(8)`	Fixed-length bit string
`bit varying(n)`	`bit varying(64)`, `varbit(64)`	Variable-length bit string
`character(n)`	`character(10)`, `char(10)`	Fixed-length character string
`character varying(n)`	`character varying(255)`, `varchar(255)`	Variable-length character string
`numeric(p,s)`	`numeric(10,2)`, `decimal(10,2)`	Exact numeric with precision and scale
`numeric(p)`	`numeric(10)`	Exact numeric with precision
`time(p)`	`time(6)`	Time of day with precision
`time with time zone`	`timetz`	Time with timezone
`time(p) with time zone`	`time(6) with time zone`	Time with timezone and precision
`time without time zone`		Time without timezone (default)
`timestamp(p)`	`timestamp(6)`	Date and time with precision
`timestamp with time zone`	`timestamptz`	Timestamp with timezone
`timestamp(p) with time zone`	`timestamp(6) with time zone`	Timestamp with timezone and precision
`timestamp without time zone`		Timestamp without timezone (default)
`vector(n)`	`vector(1536)`	pgvector extension for embeddings

Array Types

Any type can be made into an array by appending []:

columns:
  - name: tags
    type: text[]
  - name: readings
    type: double precision[]

Type Aliases

SchemaHero automatically normalizes these aliases:

You Write	SchemaHero Uses
`int8`	`bigint`
`serial8`	`bigserial`
`bool`	`boolean`
`float8`	`double precision`
`int`, `int4`	`integer`
`float4`	`real`
`int2`	`smallint`
`serial2`	`smallserial`
`serial4`	`serial`
`varbit`	`bit varying`
`char`	`character`
`varchar`	`character varying`
`decimal`	`numeric`
`timetz`	`time with time zone`
`timestamptz`	`timestamp with time zone`

Time-Series Recommendations

For hypertables, these types are recommended for time columns:

Type	Use Case
`timestamptz`	Recommended - handles timezones correctly
`timestamp`	When timezone handling is done in application
`bigint`	Unix epoch timestamps (milliseconds or microseconds)
`date`	Daily granularity only

Time-Series Example

columns:
  - name: time
    type: timestamptz
    constraints:
      notNull: true
  - name: device_id
    type: uuid
    constraints:
      notNull: true
  - name: temperature
    type: double precision
  - name: humidity
    type: double precision
  - name: location
    type: point
  - name: metadata
    type: jsonb
  - name: readings
    type: double precision[]

Tables Constraints