Content semantics
Content Semantics provides AI-powered content discovery for your Contentful spaces. It enables:
- Search — Find entries by meaning rather than exact keywords.
- Recommendations — Get entries that are semantically similar to a given set of entries.
- Duplicates — Detect potential duplicate entries based on semantic similarity.
- Reference suggestions — Get suggested entries to reference in a given field.
- Search index management — Create, list, and delete search indices that power the above features.
- Settings — Retrieve the supported locales.
Result limits
Search, recommendations, duplicates, and reference-suggestions endpoints do not support pagination or a limit parameter. Each request returns at most 10 entries.
Semantic search
Finds entries by meaning rather than exact keyword matching. The query is compared against the indexed content of your entries, so results include entries that are conceptually relevant even when they don’t contain the exact search terms.
Semantic recommendations
Returns entries that are semantically similar to a given set of entries.
Semantic duplicates
Identifies entries whose content is semantically similar to a given entry, surfacing potential duplicates even when the wording differs. Use this for content hygiene — for example, detecting blog posts or product descriptions that cover the same topic.
Semantic reference suggestions
Get semantic reference suggestions
Suggests entries that could be referenced from a specific reference field on a given entry. Suggestions are ranked by semantic relevance to the source entry, making it easier to discover and link related content.
Create semantic search index
Creates a search index for a space and locale. A search index must exist before you can use the discovery endpoints (search, recommendations, duplicates, and reference suggestions). Each index is scoped to the master environment of a single space and the selected locale. There can only be one index per space.
List semantic search index
Returns all search indices across spaces within the organization. Use the optional status query parameter to filter by index status (for example, ACTIVE to see only indices that are ready to use).
Get semantic search index
Returns the details of a single search index, including its current status. Use this to check whether an index is still being provisioned (PENDING), ready to use (ACTIVE), or being removed (DELETING).
Delete semantic search index
Permanently deletes a search index. This action is irreversible — once deleted, the index must be recreated to re-enable semantic features for that space and locale.
Semantic environment search index
Get all environment search indices
Returns search indices scoped to a specific environment. Use status=ACTIVE to check whether an active index exists before calling search, recommendations, duplicates, or reference suggestions.
Semantic settings
Returns the semantic settings for an organization, including the locale prefixes that are supported for semantic indexing. Use this to check which locales can be indexed before creating a search index.