v1.9: Add new retrieveVectors and _vectors behaviour

.code-samples.meilisearch.yaml

@@ -1287,3 +1287,13 @@ analytics_event_bind_event_1: |-
       "objectId": "0",
       "position": 0
     }'
+search_parameter_reference_retrieve_vectors_1: |-
+  curl -X POST 'localhost:7700/indexes/INDEX_NAME/search' \
+    -H 'content-type: application/json' \
+    --data-binary '{ 
+      "q": "kitchen utensils",
+      "retrieveVectors": true,
+      "hybrid": {
+        "embedder": "default"
+      }
+    }'

reference/api/documents.mdx

@@ -248,6 +248,45 @@ An array of documents. Each document is represented as a JSON object.
 ]
 ```
 
+#### `_vectors`
+
+`_vectors` is a special document attribute containing vector embeddings for AI-powered search. It accepts different types of values.
+
+`_vectors` may be an array of numbers:
+
+```json
+[
+  {
+    "id": 452,
+    "title": "Female Trouble",
+    "overview": "Delinquent high school student Dawn Davenport runs away from home and embarks upon a life of crime.",
+    "_vectors": [0.0, 0.1, 0.003, 0.75]
+  },
+]
+```
+
+`_vectors` may also be an object with two fields, `embeddings` and `regenerate`:
+
+```json
+[
+  {
+    "id": 452,
+    "title": "Female Trouble",
+    "overview": "Delinquent high school student Dawn Davenport runs away from home and embarks upon a life of crime.",
+    "_vectors": {
+      "embeddings": [],
+      "regenerate": true
+    }
+  }
+]
+```
+
+`embeddings` must be an array of numbers.
+
+`regenerate` must be a Boolean value. If `regenerate` is `true`, the supplied embeddings will be overwritten the next time Meilisearch automatically generates document embeddings. If `regenerate` is `false`, the document's embeddings will be automatically replaced by an embedder.
+
+If `_vectors` is empty or `null`, Meilisearch treats the document as not having any embeddings. This document is then ignored during AI-powered searches.
+
 ### Example
 
 <CodeSamples id="add_or_replace_documents_1" />

reference/api/search.mdx

@@ -1105,6 +1105,10 @@ Meilisearch will return an error if you use `hybrid` before activating your inst
 
 ### Vector (experimental)
 
+**Parameter**: `vector`<br />
+**Expected value**: an array of numbers<br />
+**Default value**: `null`
+
 Use a custom vector to perform a search query. Must be an array of numbers corresponding to the dimensions of the custom vector.
 
 `vector` is mandatory when performing searches with `userProvided` embedders. You may also use `vector` to override an embedder's automatic vector generation.
@@ -1118,3 +1122,17 @@ Use a custom vector to perform a search query. Must be an array of numbers corre
 <Capsule intent="warning">
 Meilisearch will return an error if you use `vector` before activating your instance's `vectorStore` and [configuring a custom embedder](/reference/api/settings#embedders-experimental).
 </Capsule>
+
+### Display `_vectors` in response (experimental)
+
+**Parameter**: `retrieveVectors`<br />
+**Expected value**: `true` or `false`<br />
+**Default value**: `false`
+
+Return document vector data in AI-powered searches.
+
+This parameter has no effect on queries that do not include `vector` or `hybrid`.
+
+#### Example
+
+<CodeSamples id="search_parameter_reference_retrieve_vectors_1" />