Filter

The Filter node processes an array of items and returns only the ones that match your conditions. Use it to narrow down data before passing it to downstream nodes: process only relevant records, remove duplicates, or extract items meeting specific criteria.

When to use Filter vs. If-Else

Node	Works on	Purpose
Filter	Arrays of items	Keep/remove items from a list based on conditions
If-Else	Single values	Branch the workflow based on a condition

Filter = "From these 100 leads, keep only the ones with a score above 80" If-Else = "If this lead's score is above 80, take path A; otherwise path B"

Configuration

Field	Type	Required	Description
Input array	FX formula	Yes	The array to filter, e.g., `{{find_all.records}}` or `{{http_request.body.items}}`
Condition	Condition builder	Yes	Filter criteria: items matching this stay, others are removed
Mode	Select	No	`Keep matching` (default) or `Remove matching`

Condition operators

Operator	Works with	Example
equals	Text, Number	`status` equals `active`
not equals	Text, Number	`status` not equals `archived`
contains	Text	`email` contains `@company.com`
does not contain	Text	`name` does not contain `test`
greater than	Number	`score` greater than `80`
less than	Number	`amount` less than `1000`
is empty	Any	`phone` is empty
is not empty	Any	`email` is not empty
starts with	Text	`url` starts with `https://`
ends with	Text	`email` ends with `.com`
matches regex	Text	`phone` matches `^\+1.*`

Multiple conditions

Add multiple conditions. Choose the logic:

Logic	Meaning
AND	All conditions must be true (default)
OR	At least one condition must be true

Output variables

{{filter.result}}         → array of items that passed the filter
{{filter.count}}          → number of items that passed
{{filter.removed}}        → number of items that were filtered out
{{filter.total}}          → original array length

Common patterns

Filter high-value leads from enrichment

Find All (all leads) → Filter (score > 80 AND status = "new") →
  For Each → Send Email (personalized outreach)

Remove already-processed items

HTTP Request (get all orders) → Filter (processed_at is empty) →
  For Each → Process order → Update Record (set processed_at)

Extract items with missing data

Find All (contacts) → Filter (email is empty OR phone is empty) →
  Google Sheets (export for manual cleanup)

Clean webhook data

Webhook (batch payload) → Filter (remove items where amount = 0) →
  For Each → Create Record

Filter + For Each pattern

Filter and For Each are commonly used together:

Filter narrows the array to relevant items
For Each processes each remaining item individually

Find All (1000 contacts) → Filter (unsubscribed = false AND last_contact > 30 days) →
  Result: 150 contacts → For Each → GPT Writer (personalized email) → Send Email

Without the Filter, you'd process all 1,000 contacts (wasting AI credits). With it, you only process the 150 that matter.

Tip

Always filter before expensive operations (AI, enrichment, API calls). Processing 100 filtered items is much cheaper than processing 1,000 items and discarding 900 of the results.

Note

Filter processes the entire array in memory. For very large arrays (10,000+ items), consider using a database query with filters instead (Find All with filter conditions) for better performance.