> For the complete documentation index, see [llms.txt](https://docs.feedspring.com/introduction/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.feedspring.com/introduction/graphql-api/querying-feeds.md).

# Querying Feeds

How the feed query works: feed types, inline fragments, collections under nodes, nullability and item limits.

Every request uses the same `feed` query. What comes back depends on the source connected to the Feed ID.

## Feed types

The `feed` field returns the `FeedData` union. Its concrete type depends on the source connected to the Feed ID:

| Source         | GraphQL type            |
| -------------- | ----------------------- |
| Instagram      | `InstagramFeedData`     |
| TikTok         | `TikTokFeedData`        |
| Dribbble       | `DribbbleFeedData`      |
| Google Reviews | `GoogleReviewsFeedData` |

Select `__typename` to identify the returned source, then use an inline fragment to select source-specific fields:

```graphql
query FeedType($publicKey: String!) {
  feed(publicKey: $publicKey) {
    __typename

    ... on InstagramFeedData {
      posts {
        nodes {
          id
        }
      }
    }

    ... on TikTokFeedData {
      videos {
        nodes {
          id
        }
      }
    }

    ... on DribbbleFeedData {
      shots {
        nodes {
          id
        }
      }
    }

    ... on GoogleReviewsFeedData {
      reviews {
        nodes {
          id
        }
      }
    }
  }
}
```

## Queries for each source

Each feed page has a complete query and notes on every field:

* [Instagram](/introduction/feeds/instagram.md#graphql-api) — `InstagramFeedData`, items in `posts.nodes`
* [Google Reviews](/introduction/feeds/google-reviews.md#graphql-api) — `GoogleReviewsFeedData`, items in `reviews.nodes`
* [TikTok](/introduction/feeds/tiktok.md#graphql-api) — `TikTokFeedData`, items in `videos.nodes`
* [Dribbble](/introduction/feeds/dribbble.md#graphql-api) — `DribbbleFeedData`, items in `shots.nodes`

## Scalars and nullability

The API uses these GraphQL scalars:

| Scalar  | Description                                                |
| ------- | ---------------------------------------------------------- |
| `ID`    | Identifier serialized as a string.                         |
| `Time`  | ISO 8601 date and time value.                              |
| `Int64` | 64-bit integer used for engagement and aggregate counters. |

Fields marked with `!` in the schema are non-null. Image objects and some source fields are nullable because providers do not always supply them.

Connection fields `posts`, `videos`, `shots`, and `reviews` are always present. Their `nodes` arrays return `[]` instead of `null` when no items are available. Nested collection fields such as `children` and `tags` also always return an array.

## Filtering and item limits

Provider collections use connection-style objects, with the current items available under `nodes`. The API returns the items currently configured and available for the feed. The `feed` query does not accept pagination, `limit`, or `skip` arguments.

Use the FeedSpring dashboard to configure source filters and the number of synced items. If one application needs fewer items, select them after receiving the response:

```javascript
const visiblePosts = result.data.feed.posts.nodes.slice(0, 6)
```

## Next steps

* [Images](/introduction/graphql-api/images.md)
* [Errors](/introduction/graphql-api/errors.md)
* [Limits & Caching](/introduction/graphql-api/limits-and-caching.md)