diff --git a/docs/content/modules/ROOT/pages/hybrid-queries.adoc b/docs/content/modules/ROOT/pages/hybrid-queries.adoc
index 9e053ac..b073552 100644
--- a/docs/content/modules/ROOT/pages/hybrid-queries.adoc
+++ b/docs/content/modules/ROOT/pages/hybrid-queries.adoc
@@ -523,6 +523,65 @@ Filter multiFilter = Filter.and(
 );
 ----
 
+== Multi-Vector Queries
+
+When your documents have multiple vector fields (e.g. text embeddings and image embeddings), use `MultiVectorQuery` to search across all of them simultaneously:
+
+[source,java]
+----
+import com.redis.vl.query.MultiVectorQuery;
+import com.redis.vl.query.Vector;
+
+Vector textVec = Vector.builder()
+    .vector(textEmbedding)
+    .fieldName("text_embedding")
+    .weight(0.7)
+    .build();
+
+Vector imageVec = Vector.builder()
+    .vector(imageEmbedding)
+    .fieldName("image_embedding")
+    .weight(0.3)
+    .build();
+
+MultiVectorQuery query = MultiVectorQuery.builder()
+    .vectors(textVec, imageVec)
+    .returnFields("title", "category")
+    .numResults(10)
+    .build();
+
+List<Map<String, Object>> results = index.query(query);
+----
+
+Results are scored using a weighted combination of the individual vector similarities.
+
+=== Per-Vector Distance Thresholds
+
+Each vector can have its own `maxDistance` to control how similar results must be for that particular field. Values range from 0.0 (exact match only) to 2.0 (include everything):
+
+[source,java]
+----
+// Tight threshold on text, loose on image
+Vector textVec = Vector.builder()
+    .vector(textEmbedding)
+    .fieldName("text_embedding")
+    .maxDistance(0.3)   // only very similar text
+    .build();
+
+Vector imageVec = Vector.builder()
+    .vector(imageEmbedding)
+    .fieldName("image_embedding")
+    .maxDistance(1.5)   // broader image matches
+    .build();
+
+MultiVectorQuery query = MultiVectorQuery.builder()
+    .vectors(textVec, imageVec)
+    .numResults(10)
+    .build();
+----
+
+When `maxDistance` is not set, it defaults to 2.0 (no distance filtering).
+
 == Tips for Effective Hybrid Queries
 
 . *Use Pre-Filtering* - More efficient than post-filtering
diff --git a/docs/content/modules/ROOT/pages/message-history.adoc b/docs/content/modules/ROOT/pages/message-history.adoc
index 274160b..c4626d6 100644
--- a/docs/content/modules/ROOT/pages/message-history.adoc
+++ b/docs/content/modules/ROOT/pages/message-history.adoc
@@ -18,7 +18,9 @@ MessageHistory chatHistory = new MessageHistory("student tutor", jedis);
 ----
 
 To align with common LLM APIs, Redis stores messages with `role` and `content` fields.
-The supported roles are "system", "user" and "llm".
+The supported roles are defined by the `ChatRole` enum: `user`, `assistant`, `system`, and `tool`.
+
+NOTE: The role `llm` is accepted for backward compatibility but is deprecated. Use `assistant` instead.
 
 You can store messages one at a time or all at once.
 
@@ -82,6 +84,19 @@ Output:
 {role=llm, content=England is larger in land area than Portugal by about 15000 square miles.}
 ----
 
+== Counting messages
+
+You can count the number of messages stored for the current session, or for a specific session tag:
+
+[source,java]
+----
+// Count messages in the default session
+long messageCount = chatHistory.count();
+
+// Count messages in a specific session
+long sessionCount = chatHistory.count("student two");
+----
+
 == Managing multiple users and conversations
 
 For applications that need to handle multiple conversations concurrently, Redis supports tagging messages to keep conversations separated.