Object
A table. Contacts, Companies, Deals, or something only your workspace has.
Property
A column on an object. Has a type, which decides what its values look like.
Record
A row in an object. One contact, one deal.
Field
One value. A record’s answer for one property.
Fields are keyed by property key
A record’sfields is a flat map keyed by each property’s key, a stable, human-readable identifier like email or annual_revenue:
key, its display name (case-insensitive), or its id, so you can post the shape your own system already has. Reads always come back keyed by key.
Every record in a list has the same
fields keys. A property with no value
is null, never missing, so you can read record.fields.email without
guarding for its absence.Discover the schema before you write
Fetch the object to learn which properties exist and what type each one is:type is what decides the value you send. Getting it wrong is the most common integration bug: a select wants one of its declared options, a relation wants a record id, a number wants a number rather than a numeric string.
Property types
The types you’ll meet most, and what a value looks like on the wire:
The full set is larger (
array, notes, whiteboard, spreadsheet, presentation, channel, conversation, and several internal ones), but those back in-app surfaces rather than integration data.
Computed properties are read-only
Aformula or rollup property has no stored value. It’s recalculated every time you read the record, from the other fields it references. So:
- You can read it like any other field.
- Sending a value for one is ignored. It will be recomputed on the next read.
- Its value reflects the moment you asked, not the moment the record was written.
Property versions
Each property carries aversion that increments when its definition changes materially. You don’t need to send or track it. It exists so the search index knows when a column’s stored values are stale. It’s visible on the property because it’s part of the honest shape, not because integrations need it.
What’s next
Objects
Read the schema, objects and their properties.
Records
Read and create the rows.