On this page
Return to top

Filter

Type key: filter
Group: Data
Category: data
Tier: CORE

The Filter node removes items from an array based on one or more conditions. It is a declarative, non-destructive way to narrow down datasets without writing code.

Handles

Handle Direction Description
input Input Receives the data array to filter
output Output Emits the filtered array

Configuration

Field Type Default Description
conditions array [] Array of condition objects
logic enum AND How conditions are combined: AND or OR

Condition Object

JSON 
{
    "type": "equals",
    "data": { "field": "data.status", "value": "active" },
    "logic": "AND"
}
Property Description
type Comparison operator (slug format)
data.field Dot-notation path to the field in each array item
data.value Value to compare against
logic Logic connector with next condition: AND or OR

Smart Type Inference

The Filter node automatically detects data types from upstream JSON values and adapts the UI accordingly:

Detected type How detected Operators shown Value input
Number typeof value === 'number' in JSON =, !=, >, <, >=, <=, in, not_in, is_null, is_not_null Text input
String typeof value === 'string' in JSON =, !=, contains, not_contains, starts_with, ends_with, in, not_in, is_null, is_not_null Select from upstream values or text input (toggle with ⚡)
Boolean typeof value === 'boolean' in JSON =, !=, is_null, is_not_null, is_true, is_false True/False dropdown
Date Field name ends with _at or _date is_exactly, is_not, is_after, is_before, is_in_the_past, is_today, is_within_X_days Date picker

When connected to a Data Model node, field types are inferred from the database column types via the model schema.

Value Picker (⚡)

For string fields, the value input shows a select dropdown with distinct values extracted from upstream JSON data. Click the ⚡ icon to toggle between:

  • Select mode: pick from available values
  • Manual mode: enter free text or dynamic expressions like {{variable}}

Supported Operators

Operator Description
equals Strict equality
not_equals Not equal
greater_than Numeric >
less_than Numeric <
greater_than_or_equal Numeric >=
less_than_or_equal Numeric <=
contains String contains substring
not_contains String does not contain
starts_with String starts with
ends_with String ends with
in Value is in a comma-separated list
not_in Value is not in list
is_null Value is null
is_not_null Value is not null
is_empty Null, empty string, or empty array
is_not_empty Not null and not empty
is_in_the_past Date is in the past
is_in_the_future Date is in the future
is_today Date is today
is_within_days Date within X days from now
is_within_months Date within X months from now
is_older_than_days Date older than X days
is_older_than_months Date older than X months

Logic Modes

AND (default)

All conditions must be satisfied for an item to pass:

TEXT 
Field: data.status  = active
AND
Field: data.score   > 50

Only items where both status = active AND score > 50 are included.

OR

At least one condition must be satisfied:

TEXT 
Field: data.status  = active
OR
Field: data.status  = trial

Items where status is either active OR trial are included.

Input Detection

The Filter node automatically detects the array to filter:

  1. If input has a data key with an array → filters data
  2. If input itself is a flat array → filters it directly

The structure is preserved in the output (i.e., if the input had a data key, the output also has a data key).

Field Path Resolution

Field paths use dot notation from the root JSON (e.g., data.status, data.lifetime_value). When evaluating conditions against individual items already extracted from a collection, the Filter node progressively strips leading path segments to find the correct value.

Output

Given input:

JSON 
{
    "data": [
        { "id": 1, "status": "active", "score": 80 },
        { "id": 2, "status": "inactive", "score": 90 },
        { "id": 3, "status": "active", "score": 30 }
    ]
}

With conditions: data.status = active AND data.score > 50

Output:

JSON 
{
    "data": [{ "id": 1, "status": "active", "score": 80 }],
    "count": 1
}

Preset Example

Filter node — active customers example Filter node — active customers example

::: info 💾 Try this example Download the workflow JSON — open an empty workflow and import the file. :::

Examples

Filter Active High-Value Customers

YAML 
Logic: AND
Conditions:
    - data.status         equals          active
    - data.lifetime_value greater_than    500

Filter Items with Missing Email

YAML 
Logic: AND
Conditions:
    - data.email  is_not_empty

Filter by Multiple Statuses (OR)

YAML 
Logic: OR
Conditions:
    - data.status  equals  pending
    - data.status  equals  processing

Notes

  • If no conditions are configured, all input data passes through unchanged
  • The Filter node does not modify the items it passes — only includes/excludes them
  • Operators shown in the UI adapt based on the detected field data type
  • For transforming item fields, use the Transform node
  • For a single conditional branch (not array filtering), use the If node