Caching and Fresh Data

Trakt API responses generally contain two different kinds of data:

• Metadata, such as titles, overviews, taglines, images, translations, ratings, IDs, genres, and updated_at.
• Activity or user-specific data, such as watched state, favorites, plays, list counts, progress, and other counts.

These two data types are cached and updated differently.

Summary endpoints are the source of truth for metadata

For the freshest metadata, use the dedicated summary endpoints:

• GET /movies/{id}?extended=full
• GET /shows/{id}?extended=full
• GET /shows/{id}/seasons/{season}/episodes/{episode}?extended=full

These endpoints should be treated as the canonical source for current metadata fields like overview, tagline, available_translations, images, social IDs, ratings data, comments, and updated_at.

List, activity, and sync endpoints may return cached snapshots

Some endpoints return media objects nested inside a larger response. For example:

• GET /movies/anticipated
• GET /movies/watched/{period}
• GET /movies/played/{period}
• GET /movies/collected/{period}
• GET /movies/favorited/{period}
• GET /sync/watched/{type}
• GET /sync/collection/{type}
• user history, collection, ratings, favorites, and watchlist endpoints

These endpoints are optimized for the list, activity, sync, or user-specific data they return. Any nested movie, show, season, or episode object should be treated as a convenience snapshot for matching and display, not as the freshest metadata record.

Even when extended=full is supported, the embedded object might not be updated at the exact same time as the corresponding summary endpoint.

Recommended client pattern

Use activity and sync endpoints to get IDs and user-specific state.

Then, when you need current metadata, hydrate those items separately from the summary endpoints.