> ## Documentation Index
> Fetch the complete documentation index at: https://www.meilisearch.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Precise filtering with array relationships

> Use foreign filters to precisely filter documents with array relationships using AND logic across items.

When a relationship contains an array of related documents, foreign filters enable precise filtering with AND logic across items, unlike normal filters which use OR logic.

## Array filter logic

Normal filters use **OR logic** across array items:

* Find films where `actors.country = "France" OR actors.age > 35`
* Returns films with ANY actor who is French OR over 35 (possibly different actors)

Foreign filters use **AND logic** across array items:

* Find films where `_foreign(actors, country = "France" AND age > 35)`
* Returns films with an actor who is BOTH French AND over 35 (same actor meets both conditions)

## Why this matters

Precise filtering is essential when you need to find documents where a related item meets **multiple conditions simultaneously**. Without this capability, you'd need to denormalize data or post-process results in your application code.

## Film cast example

Consider a film with three actors:

```json theme={null}
{
  "id": "film_1",
  "title": "Ocean's Eleven",
  "actor_ids": ["actor_1", "actor_2", "actor_3"]
}
```

Where the actors are:

```json theme={null}
[
  {"id": "actor_1", "name": "Actor A", "country": "USA", "age": 45},
  {"id": "actor_2", "name": "Actor B", "country": "France", "age": 38},
  {"id": "actor_3", "name": "Actor C", "country": "France", "age": 30}
]
```

### Query 1: Normal filter (OR logic)

<CodeGroup>
  ```bash cURL theme={null}
  curl \
    -X POST 'MEILISEARCH_URL/indexes/films/search' \
    -H 'Content-Type: application/json' \
    --data-binary '{
      "filter": "actors.country = \"France\" OR actors.age > 35"
    }'
  ```
</CodeGroup>

**Result:** Film is included because:

* Actor B is French (country = "France") ✓
* Actor A is over 35 (age > 35) ✓

Both conditions are met, but by **different actors**.

### Query 2: Foreign filter (AND logic)

<CodeGroup>
  ```bash cURL theme={null}
  curl \
    -X POST 'MEILISEARCH_URL/indexes/films/search' \
    -H 'Content-Type: application/json' \
    --data-binary '{
      "filter": "_foreign(actors, country = \"France\" AND age > 35)"
    }'
  ```
</CodeGroup>

**Result:** Film is included because:

* Actor B is BOTH French AND over 35 ✓

A **single actor** meets both conditions.

## Real-world examples

### Find products with multiple required certifications

```json theme={null}
{
  "id": "product_1",
  "name": "Industrial Motor",
  "certification_ids": ["cert_1", "cert_2", "cert_3"]
}
```

Find products that have **both** ISO 9001 and CE certifications issued after 2020:

<CodeGroup>
  ```bash cURL theme={null}
  curl \
    -X POST 'MEILISEARCH_URL/indexes/products/search' \
    -H 'Content-Type: application/json' \
    --data-binary '{
      "q": "motor",
      "filter": "_foreign(certifications, (standard = \"ISO 9001\" OR standard = \"CE\") AND issued_year >= 2020)"
    }'
  ```
</CodeGroup>

### Find articles with specific tag combinations

Find articles that have **both** the "security" AND "encryption" tags (same article, not different articles):

<CodeGroup>
  ```bash cURL theme={null}
  curl \
    -X POST 'MEILISEARCH_URL/indexes/articles/search' \
    -H 'Content-Type: application/json' \
    --data-binary '{
      "q": "cryptography",
      "filter": "_foreign(tags, name = \"security\" AND name = \"encryption\")"
    }'
  ```
</CodeGroup>

Wait—this won't work as expected because a single tag can't have multiple names. For this use case, you'd want to check if an article has both tags, which would require a different approach (checking two separate arrays or normalizing differently).

### Find employees with specific skill levels

Find employees with **both** Python AND JavaScript at advanced level:

<CodeGroup>
  ```bash cURL theme={null}
  curl \
    -X POST 'MEILISEARCH_URL/indexes/employees/search' \
    -H 'Content-Type: application/json' \
    --data-binary '{
      "q": "fullstack",
      "filter": "_foreign(skills, (language = \"Python\" AND proficiency = \"advanced\") OR (language = \"JavaScript\" AND proficiency = \"advanced\"))"
    }'
  ```
</CodeGroup>

## When to use precise filtering

Use foreign filters with AND logic when:

* A document has an array of related items
* You need to find items where **a single related item** meets multiple conditions
* You're filtering on nested properties of array items

Use normal filters with OR logic when:

* You're checking if **any item** in an array matches your criteria
* You need broader matching across multiple array items

## Next steps

<CardGroup cols={2}>
  <Card title="Foreign filters overview" href="/capabilities/filtering_sorting_faceting/advanced/filtering_by_joined_data">
    Learn about foreign filters and filtering by joined data
  </Card>

  <Card title="Define relationships" href="/capabilities/indexing/joins/define_index_relationships">
    Set up the relationships this filtering depends on
  </Card>
</CardGroup>