--- title: "Search a managed index" method: POST path: "/v1/rag/managed_indexes/{index_name}/search" tags: ["beta.rag.managed_indexes"] --- # Search a managed index `POST /v1/rag/managed_indexes/{index_name}/search` Runs an approximate-nearest-neighbor vector search over the index and returns matching chunks, closest-first. ## Path parameters - `index_name` string, required ## Request body - SearchRequest - `retriever` union, required - KeywordRetriever — Retrieve document chunks with similar keywords to the given query. - `top_k` integer - `type` 'keyword' - `query` string, required - `filter` union - And — Conjunction: rows must match every sub-filter. - `type` 'and' - `matches` union[], required - union - … - Or — Disjunction: rows must match at least one sub-filter. - `type` 'or' - `matches` union[], required - union - … - Not — Negate a filter: rows must not match. - `type` 'not' - `match` union, required - And — Conjunction: rows must match every sub-filter. - … - Or — Disjunction: rows must match at least one sub-filter. - … - Not — recursive - Equal — Field equals a scalar value. - … - Range — Check that the given field falls in a range. Requires at least one operator. - … - In — Field is one of a set of values. - … - Equal — Field equals a scalar value. - `type` 'equal' - `field` string, required - `value` union, required - integer - number - string, date-time - string - Range — Check that the given field falls in a range. Requires at least one operator. - `type` 'range' - `field` string, required - `gt` union - integer - number - string, date-time - `gte` union - integer - number - string, date-time - `lt` union - integer - number - string, date-time - `lte` union - integer - number - string, date-time - In — Field is one of a set of values. - `type` 'in' - `field` string, required - `values` union[], required - union - … - `field` string, nullable - NearestNeighbourRetriever — Retrieve document chunks whose embeddings are closest to the query's. - `top_k` integer - `type` 'nearest_neighbour' - `query` string, nullable - `query_embedding` number[], nullable - `field` string, nullable - `filter` union - And — Conjunction: rows must match every sub-filter. - `type` 'and' - `matches` union[], required - union - … - Or — Disjunction: rows must match at least one sub-filter. - `type` 'or' - `matches` union[], required - union - … - Not — Negate a filter: rows must not match. - `type` 'not' - `match` union, required - And — Conjunction: rows must match every sub-filter. - … - Or — Disjunction: rows must match at least one sub-filter. - … - Not — recursive - Equal — Field equals a scalar value. - … - Range — Check that the given field falls in a range. Requires at least one operator. - … - In — Field is one of a set of values. - … - Equal — Field equals a scalar value. - `type` 'equal' - `field` string, required - `value` union, required - integer - number - string, date-time - string - Range — Check that the given field falls in a range. Requires at least one operator. - `type` 'range' - `field` string, required - `gt` union - integer - number - string, date-time - `gte` union - integer - number - string, date-time - `lt` union - integer - number - string, date-time - `lte` union - integer - number - string, date-time - In — Field is one of a set of values. - `type` 'in' - `field` string, required - `values` union[], required - union - … - RRFRetriever — Fuse child retriever rankings via reciprocal rank fusion. - `top_k` integer - `type` 'rrf' - `retrievers` union[], required - union - KeywordRetriever — Retrieve document chunks with similar keywords to the given query. - `top_k` integer - `type` 'keyword' - `query` string, required - `filter` union - … - `field` string, nullable - NearestNeighbourRetriever — Retrieve document chunks whose embeddings are closest to the query's. - `top_k` integer - `type` 'nearest_neighbour' - `query` string, nullable - `query_embedding` number[], nullable - `field` string, nullable - `filter` union - … - `weights` number[], nullable - `rank_constant` integer - `filter` union - And — Conjunction: rows must match every sub-filter. - `type` 'and' - `matches` union[], required - union - … - Or — Disjunction: rows must match at least one sub-filter. - `type` 'or' - `matches` union[], required - union - … - Not — Negate a filter: rows must not match. - `type` 'not' - `match` union, required - And — Conjunction: rows must match every sub-filter. - … - Or — Disjunction: rows must match at least one sub-filter. - … - Not — recursive - Equal — Field equals a scalar value. - … - Range — Check that the given field falls in a range. Requires at least one operator. - … - In — Field is one of a set of values. - … - Equal — Field equals a scalar value. - `type` 'equal' - `field` string, required - `value` union, required - integer - number - string, date-time - string - Range — Check that the given field falls in a range. Requires at least one operator. - `type` 'range' - `field` string, required - `gt` union - integer - number - string, date-time - `gte` union - integer - number - string, date-time - `lt` union - integer - number - string, date-time - `lte` union - integer - number - string, date-time - In — Field is one of a set of values. - `type` 'in' - `field` string, required - `values` union[], required - union - … ## Response `200` Successful Response - SearchResponse - `hits` SearchHitResponse[], required - `chunk` SearchChunkResponse, required — Wire chunk: atom fields plus any custom fields (via ``extra``). - `id` string, required - `source_id` string, required - `locator` string, required - `start_offset` integer, nullable, required - `end_offset` integer, nullable, required - `chunk_type` string, required - `parent_ref` string, nullable - `content` string, required - `metadata` object - `score` number, required - `distance` number, nullable - `index_name` string, required ## Other responses - `400` — Invalid search query - `404` — Index not found - `409` — Index is not ready for search - `422` — Validation Error - `500` — Search execution failed ## Changes - **2026-09-23** `fa73befe1c61` — 1 info - endpoint added [Change history](https://skmtc.dev/mistral/apis/mistral-ai-api/changes/v1/rag/managed_indexes/:index_name/search/post.md) --- [API](https://skmtc.dev/mistral/apis/mistral-ai-api.md) · [All operations](https://skmtc.dev/mistral/apis/mistral-ai-api/llms.txt) · [OpenAPI document](https://skmtc.dev/mistral/apis/mistral-ai-api/revisions/fa73befe1c61?raw)