diff --git a/docs/reference/api-reference.md b/docs/reference/api-reference.md
index baa9031c9..055906a5a 100644
--- a/docs/reference/api-reference.md
+++ b/docs/reference/api-reference.md
@@ -7878,7 +7878,7 @@ client.inference.putCustom({ task_type, custom_inference_id, service, service_se
 - **`task_type` (Enum("text_embedding" \| "sparse_embedding" \| "rerank" \| "completion"))**: The type of the inference task that the model will perform.
 - **`custom_inference_id` (string)**: The unique identifier of the inference endpoint.
 - **`service` (Enum("custom"))**: The type of service supported for the specified task type. In this case, `custom`.
-- **`service_settings` ({ headers, input_type, query_parameters, request, response, secret_parameters, url })**: Settings used to install the inference model.
+- **`service_settings` ({ batch_size, headers, input_type, query_parameters, request, response, secret_parameters, url })**: Settings used to install the inference model.
 These settings are specific to the `custom` service.
 - **`chunking_settings` (Optional, { max_chunk_size, overlap, sentence_overlap, separator_group, separators, strategy })**: The chunking configuration object.
 - **`task_settings` (Optional, { parameters })**: Settings to configure the inference task.
diff --git a/src/api/types.ts b/src/api/types.ts
index 4caf5ee8d..cea5eb911 100644
--- a/src/api/types.ts
+++ b/src/api/types.ts
@@ -7420,7 +7420,7 @@ export interface AnalysisKeywordTokenizer extends AnalysisTokenizerBase {
 
 export interface AnalysisKuromojiAnalyzer {
   type: 'kuromoji'
-  mode: AnalysisKuromojiTokenizationMode
+  mode?: AnalysisKuromojiTokenizationMode
   user_dictionary?: string
 }
 
@@ -22979,7 +22979,11 @@ export interface InferenceCustomResponseParams {
 }
 
 export interface InferenceCustomServiceSettings {
-  /** Specifies the HTTPS header parameters – such as `Authentication` or `Contet-Type` – that are required to access the custom service.
+  /** Specifies the batch size used for the semantic_text field. If the field is not provided, the default is 10.
+    * The batch size is the maximum number of inputs in a single request to the upstream service.
+    * The chunk within the batch are controlled by the selected chunking strategy for the semantic_text field. */
+  batch_size?: integer
+  /** Specifies the HTTP header parameters – such as `Authentication` or `Content-Type` – that are required to access the custom service.
     * For example:
     * ```
     * "headers":{