diff --git a/ml-algorithms/src/main/java/org/opensearch/ml/engine/algorithms/remote/RemoteConnectorExecutor.java b/ml-algorithms/src/main/java/org/opensearch/ml/engine/algorithms/remote/RemoteConnectorExecutor.java
index e6fa6331bc..1ccd6f1879 100644
--- a/ml-algorithms/src/main/java/org/opensearch/ml/engine/algorithms/remote/RemoteConnectorExecutor.java
+++ b/ml-algorithms/src/main/java/org/opensearch/ml/engine/algorithms/remote/RemoteConnectorExecutor.java
@@ -83,10 +83,10 @@ default void executePredict(MLInput mlInput, ActionListener<MLTaskResponse> acti
 
     /**
      * Calculate the chunk size.
-     * @param textDocsInputDataSet
+     * @param textDocsInputDataSet Input dataset in textDocsInputDataSet format.
      * @return Tuple of chunk size and step size.
      */
-    private Tuple<Integer, Integer> calculateChunkSize(TextDocsInputDataSet textDocsInputDataSet) {
+    default Tuple<Integer, Integer> calculateChunkSize(TextDocsInputDataSet textDocsInputDataSet) {
         int textDocsLength = textDocsInputDataSet.getDocs().size();
         Map<String, String> parameters = getConnector().getParameters();
         if (parameters != null && parameters.containsKey("input_docs_processed_step_size")) {
@@ -107,11 +107,15 @@ private Tuple<Integer, Integer> calculateChunkSize(TextDocsInputDataSet textDocs
                 throw new IllegalArgumentException("no predict action found");
             }
             String preProcessFunction = predictAction.get().getPreProcessFunction();
-            if (preProcessFunction != null && !MLPreProcessFunction.contains(preProcessFunction)) {
-                // user defined preprocess script, this case, the chunk size is always equals to text docs length.
+            if (preProcessFunction == null) {
+                // default preprocess case, consider this a batch.
+                return Tuple.tuple(1, textDocsLength);
+            } else if (MLPreProcessFunction.TEXT_DOCS_TO_BEDROCK_EMBEDDING_INPUT.equals(preProcessFunction)
+                || !MLPreProcessFunction.contains(preProcessFunction)) {
+                // bedrock and user defined preprocess script, the chunk size is always equals to text docs length.
                 return Tuple.tuple(textDocsLength, 1);
             }
-            // consider as batch.
+            //Other cases: non-bedrock and user defined preprocess script, consider as batch.
             return Tuple.tuple(1, textDocsLength);
         }
     }
diff --git a/ml-algorithms/src/test/java/org/opensearch/ml/engine/algorithms/remote/AwsConnectorExecutorTest.java b/ml-algorithms/src/test/java/org/opensearch/ml/engine/algorithms/remote/AwsConnectorExecutorTest.java
index 13b2ceb987..6465c99710 100644
--- a/ml-algorithms/src/test/java/org/opensearch/ml/engine/algorithms/remote/AwsConnectorExecutorTest.java
+++ b/ml-algorithms/src/test/java/org/opensearch/ml/engine/algorithms/remote/AwsConnectorExecutorTest.java
@@ -366,4 +366,75 @@ public void executePredict_TextDocsInferenceInput_withoutStepSize_userDefinedPre
         executor
             .executePredict(MLInput.builder().algorithm(FunctionName.TEXT_EMBEDDING).inputDataset(inputDataSet).build(), actionListener);
     }
+
+    @Test
+    public void executePredict_TextDocsInferenceInput_withoutStepSize_bedRockEmbeddingPreProcessFunction() {
+        ConnectorAction predictAction = ConnectorAction
+            .builder()
+            .actionType(ConnectorAction.ActionType.PREDICT)
+            .method("POST")
+            .url("http://openai.com/mock")
+            .requestBody("{\"input\": ${parameters.input}}")
+            .preProcessFunction(MLPreProcessFunction.TEXT_DOCS_TO_BEDROCK_EMBEDDING_INPUT)
+            .build();
+        Map<String, String> credential = ImmutableMap
+            .of(ACCESS_KEY_FIELD, encryptor.encrypt("test_key"), SECRET_KEY_FIELD, encryptor.encrypt("test_secret_key"));
+        Map<String, String> parameters = ImmutableMap.of(REGION_FIELD, "us-west-2", SERVICE_NAME_FIELD, "bedrock");
+        Connector connector = AwsConnector
+            .awsConnectorBuilder()
+            .name("test connector")
+            .version("1")
+            .protocol("aws_sigv4")
+            .parameters(parameters)
+            .credential(credential)
+            .actions(Arrays.asList(predictAction))
+            .build();
+        connector.decrypt((c) -> encryptor.decrypt(c));
+        AwsConnectorExecutor executor = spy(new AwsConnectorExecutor(connector));
+        Settings settings = Settings.builder().build();
+        threadContext = new ThreadContext(settings);
+        when(executor.getClient()).thenReturn(client);
+        when(client.threadPool()).thenReturn(threadPool);
+        when(threadPool.getThreadContext()).thenReturn(threadContext);
+        when(executor.getScriptService()).thenReturn(scriptService);
+
+        MLInputDataset inputDataSet = TextDocsInputDataSet.builder().docs(ImmutableList.of("input1", "input2", "input3")).build();
+        executor
+            .executePredict(MLInput.builder().algorithm(FunctionName.TEXT_EMBEDDING).inputDataset(inputDataSet).build(), actionListener);
+    }
+
+    @Test
+    public void executePredict_TextDocsInferenceInput_withoutStepSize_emptyPreprocessFunction() {
+        ConnectorAction predictAction = ConnectorAction
+            .builder()
+            .actionType(ConnectorAction.ActionType.PREDICT)
+            .method("POST")
+            .url("http://openai.com/mock")
+            .requestBody("{\"input\": ${parameters.input}}")
+            .build();
+        Map<String, String> credential = ImmutableMap
+            .of(ACCESS_KEY_FIELD, encryptor.encrypt("test_key"), SECRET_KEY_FIELD, encryptor.encrypt("test_secret_key"));
+        Map<String, String> parameters = ImmutableMap.of(REGION_FIELD, "us-west-2", SERVICE_NAME_FIELD, "bedrock");
+        Connector connector = AwsConnector
+            .awsConnectorBuilder()
+            .name("test connector")
+            .version("1")
+            .protocol("aws_sigv4")
+            .parameters(parameters)
+            .credential(credential)
+            .actions(Arrays.asList(predictAction))
+            .build();
+        connector.decrypt((c) -> encryptor.decrypt(c));
+        AwsConnectorExecutor executor = spy(new AwsConnectorExecutor(connector));
+        Settings settings = Settings.builder().build();
+        threadContext = new ThreadContext(settings);
+        when(executor.getClient()).thenReturn(client);
+        when(client.threadPool()).thenReturn(threadPool);
+        when(threadPool.getThreadContext()).thenReturn(threadContext);
+        when(executor.getScriptService()).thenReturn(scriptService);
+
+        MLInputDataset inputDataSet = TextDocsInputDataSet.builder().docs(ImmutableList.of("input1", "input2", "input3")).build();
+        executor
+            .executePredict(MLInput.builder().algorithm(FunctionName.TEXT_EMBEDDING).inputDataset(inputDataSet).build(), actionListener);
+    }
 }