diff --git a/integrations/vector-stores/elasticsearch/src/main/java/org/apache/flink/agents/integrations/vectorstores/elasticsearch/ElasticsearchVectorStore.java b/integrations/vector-stores/elasticsearch/src/main/java/org/apache/flink/agents/integrations/vectorstores/elasticsearch/ElasticsearchVectorStore.java
index 5d66e479f..4be3d6be1 100644
--- a/integrations/vector-stores/elasticsearch/src/main/java/org/apache/flink/agents/integrations/vectorstores/elasticsearch/ElasticsearchVectorStore.java
+++ b/integrations/vector-stores/elasticsearch/src/main/java/org/apache/flink/agents/integrations/vectorstores/elasticsearch/ElasticsearchVectorStore.java
@@ -112,7 +112,7 @@ public class ElasticsearchVectorStore extends BaseVectorStore
 
     /** Default vector dimensionality used when {@code dims} is not provided. */
     public static final int DEFAULT_DIMENSION = 768;
-    /** The maximum number of documents that can be retrieved in get. */
+    /** Default request size used by {@code get} when no limit is provided. */
     public static final int MAX_RESULT_WINDOW = 10000;
 
     public static final String DEFAULT_METADATA_FIELD = "_metadata";
@@ -336,23 +336,31 @@ public Map<String, Object> getStoreKwargs() {
     /**
      * Retrieve documents from the vector store.
      *
-     * <p>If ids is not provided, this method will retrieve documents according to {@code limit},
-     * {@code offset}, and {@code filter_query} in additional arguments. If {@code limit} is null,
-     * up to {@link ElasticsearchVectorStore#MAX_RESULT_WINDOW} documents are returned (an
-     * Elasticsearch ceiling).
+     * <p>When {@code ids} is non-empty, documents are retrieved directly by ID and the filter,
+     * limit, offset, and {@code filter_query} arguments are not applied.
      *
-     * <p>The unified {@code filters} DSL parameter is not yet translated to Elasticsearch's native
-     * query DSL — callers needing structured filtering should pass a raw {@code filter_query} via
-     * {@code extraArgs}. TODO: implement equality-DSL translation parallel to the Python Chroma
-     * implementation.
+     * <p>Otherwise, {@code filters} provides equality-only matching against document metadata. Each
+     * entry is translated to an Elasticsearch {@code term} query on {@code
+     * <metadataField>.<key>.keyword}, and multiple entries are combined with AND semantics. Because
+     * Elasticsearch dynamic mapping creates {@code .keyword} sub-fields only for strings, filters
+     * on non-string metadata values do not match; use a raw {@code filter_query} for those values.
+     * A raw Elasticsearch JSON query may also be supplied as {@code filter_query} in {@code
+     * extraArgs}; when both forms are present, they are combined with AND semantics.
      *
-     * @param ids The ids of the documents.
-     * @param collection The name of the collection to be retrieved. If is null, retrieve the
-     *     default collection.
-     * @param filters Unified filter DSL. Currently ignored — see method Javadoc.
-     * @param limit Maximum number of documents to return; falls back to {@link
-     *     ElasticsearchVectorStore#MAX_RESULT_WINDOW} when null.
-     * @param extraArgs Additional arguments. (offset, filter_query, etc.)
+     * <p>The {@code limit} parameter takes precedence over a {@code limit} value in {@code
+     * extraArgs}. If neither limit is provided, the request size defaults to {@link
+     * ElasticsearchVectorStore#MAX_RESULT_WINDOW} (10,000). The effective Elasticsearch
+     * result-window limit is controlled by the target index's {@code index.max_result_window}
+     * setting, which defaults to 10,000 but is configurable. Elasticsearch rejects requests when
+     * the combined {@code offset} and request size exceed that setting rather than truncating them.
+     * {@code extraArgs} may also contain an {@code offset}.
+     *
+     * @param ids The IDs of documents to retrieve directly.
+     * @param collection The collection name, or null to use the default collection.
+     * @param filters Equality-only metadata filters combined with AND semantics.
+     * @param limit Maximum number of documents to return for filtered or unfiltered searches.
+     * @param extraArgs Additional arguments, including {@code offset} and raw JSON {@code
+     *     filter_query}.
      * @return List of documents retrieved.
      */
     @Override
@@ -379,22 +387,24 @@ public List<Document> get(
     }
 
     /**
-     * Delete documents in the vector store.
+     * Delete documents from the vector store.
      *
-     * <p>If ids is not provided, this method will delete documents matched the {@code filter_query}
-     * in additional arguments. If neither {@code filter_query} nor {@code filters} is provided,
-     * this method will delete all the documents.
+     * <p>When {@code ids} is non-empty, documents are deleted directly by ID and {@code filters}
+     * and {@code filter_query} are not applied.
      *
-     * <p>The unified {@code filters} DSL parameter is not yet translated to Elasticsearch's native
-     * query DSL — callers needing structured filtering should pass a raw {@code filter_query} via
-     * {@code extraArgs}. TODO: implement equality-DSL translation parallel to the Python Chroma
-     * implementation.
+     * <p>Otherwise, {@code filters} provides equality-only matching against document metadata. Each
+     * entry is translated to an Elasticsearch {@code term} query on {@code
+     * <metadataField>.<key>.keyword}, and multiple entries are combined with AND semantics. Because
+     * Elasticsearch dynamic mapping creates {@code .keyword} sub-fields only for strings, filters
+     * on non-string metadata values do not match; use a raw {@code filter_query} for those values.
+     * A raw Elasticsearch JSON query may also be supplied as {@code filter_query} in {@code
+     * extraArgs}; when both forms are present, they are combined with AND semantics. If neither
+     * form is supplied, all documents in the collection are deleted.
      *
-     * @param ids The ids of the documents.
-     * @param collection The name of the collection the documents belong to. If is null, use the
-     *     default collection.
-     * @param filters Unified filter DSL. Currently ignored — see method Javadoc.
-     * @param extraArgs Additional arguments. (filter_query, etc.)
+     * @param ids The IDs of documents to delete directly.
+     * @param collection The collection name, or null to use the default collection.
+     * @param filters Equality-only metadata filters combined with AND semantics.
+     * @param extraArgs Additional arguments, including raw JSON {@code filter_query}.
      */
     @Override
     public void delete(
@@ -571,19 +581,26 @@ private void deleteDocuments(
      * Executes a KNN vector search using a pre-computed embedding.
      *
      * <p>The method prepares a KNN search request using the supplied {@code embedding} and merges
-     * default arguments from the store with the provided {@code args}. Optional filter queries
-     * (JSON DSL) restrict the documents the KNN search may match, so the nearest neighbours are
+     * default arguments from the store with the provided {@code args}. {@code filters} provides
+     * equality-only matching against metadata fields. Each entry targets {@code
+     * <metadataField>.<key>.keyword}; multiple entries are combined with AND semantics. Because
+     * Elasticsearch dynamic mapping creates {@code .keyword} sub-fields only for strings, filters
+     * on non-string metadata values do not match; use a raw {@code filter_query} for those values.
+     *
+     * <p>A raw Elasticsearch JSON query may also be supplied as {@code filter_query} in {@code
+     * args}. When both filter forms are present, they are combined with AND semantics. The combined
+     * filter restricts the documents the KNN search may match, so the nearest neighbours are
      * selected from among the matching documents rather than filtered out afterwards. Up to {@code
      * k} matching documents are returned even when the closest vectors overall do not match.
      *
-     * @param embedding The embedding vector to search with
-     * @param limit Maximum number of items the caller is interested in; used as a fallback for
-     *     {@code k} if not explicitly provided
-     * @param collection The index to query search. If is null, search the default index.
-     * @param args Additional arguments. Supported keys: {@code k}, {@code num_candidates}, {@code
-     *     filter_query}
-     * @return A list of matching documents, possibly empty
-     * @throws RuntimeException if the search request fails
+     * @param embedding The embedding vector to search with.
+     * @param limit Maximum number of items requested; used as a fallback for {@code k}.
+     * @param collection The collection name, or null to use the default collection.
+     * @param filters Equality-only metadata filters combined with AND semantics.
+     * @param args Additional arguments. Supported keys are {@code k}, {@code num_candidates}, and
+     *     raw JSON {@code filter_query}.
+     * @return A list of matching documents, possibly empty.
+     * @throws RuntimeException if the search request fails.
      */
     @SuppressWarnings({"rawtypes", "unchecked"})
     @Override