diff --git a/build-android-arm64-v8a.sh b/build-android-arm64-v8a.sh
index c6778380a..5599dc7c3 100755
--- a/build-android-arm64-v8a.sh
+++ b/build-android-arm64-v8a.sh
@@ -76,8 +76,42 @@ cmake -DCMAKE_TOOLCHAIN_FILE="$ANDROID_NDK/build/cmake/android.toolchain.cmake"
     -DANDROID_ABI="arm64-v8a" \
     -DANDROID_PLATFORM=android-21 ..
 
+# Please use -DANDROID_PLATFORM=android-27 if you want to use Android NNAPI
+
 # make VERBOSE=1 -j4
 make -j4
 make install/strip
 cp -fv $onnxruntime_version/jni/arm64-v8a/libonnxruntime.so install/lib
 rm -rf install/lib/pkgconfig
+
+# To run the generated binaries on Android, please use the following steps.
+#
+#
+# 1. Copy sherpa-onnx and its dependencies to Android
+#
+#   cd build-android-arm64-v8a/install/lib
+#   adb push ./lib*.so /data/local/tmp
+#   cd ../bin
+#   adb push ./sherpa-onnx /data/local/tmp
+#
+# 2. Login into Android
+#
+#   adb shell
+#   cd /data/local/tmp
+#   ./sherpa-onnx
+#
+# which shows the following error log:
+#
+#  CANNOT LINK EXECUTABLE "./sherpa-onnx": library "libsherpa-onnx-core.so" not found: needed by main executable
+#
+# Please run:
+#
+#  export LD_LIBRARY_PATH=$PWD:$LD_LIBRARY_PATH
+#
+# and then you can run:
+#
+#  ./sherpa-onnx
+#
+# It should show the help message of sherpa-onnx.
+#
+# Please use the above approach to copy model files to your phone.
diff --git a/sherpa-onnx/csrc/provider.cc b/sherpa-onnx/csrc/provider.cc
index dcf3a261d..95bc18c5f 100644
--- a/sherpa-onnx/csrc/provider.cc
+++ b/sherpa-onnx/csrc/provider.cc
@@ -22,6 +22,8 @@ Provider StringToProvider(std::string s) {
     return Provider::kCoreML;
   } else if (s == "xnnpack") {
     return Provider::kXnnpack;
+  } else if (s == "nnapi") {
+    return Provider::kNNAPI;
   } else {
     SHERPA_ONNX_LOGE("Unsupported string: %s. Fallback to cpu", s.c_str());
     return Provider::kCPU;
diff --git a/sherpa-onnx/csrc/provider.h b/sherpa-onnx/csrc/provider.h
index 05ce71654..467e5dab5 100644
--- a/sherpa-onnx/csrc/provider.h
+++ b/sherpa-onnx/csrc/provider.h
@@ -17,6 +17,7 @@ enum class Provider {
   kCUDA = 1,     // CUDAExecutionProvider
   kCoreML = 2,   // CoreMLExecutionProvider
   kXnnpack = 3,  // XnnpackExecutionProvider
+  kNNAPI = 4,    // NnapiExecutionProvider
 };
 
 /**
diff --git a/sherpa-onnx/csrc/session.cc b/sherpa-onnx/csrc/session.cc
index 827c4e05b..f4cb760ef 100644
--- a/sherpa-onnx/csrc/session.cc
+++ b/sherpa-onnx/csrc/session.cc
@@ -15,6 +15,10 @@
 #include "coreml_provider_factory.h"  // NOLINT
 #endif
 
+#if __ANDROID_API__ >= 27
+#include "nnapi_provider_factory.h"
+#endif
+
 namespace sherpa_onnx {
 
 static Ort::SessionOptions GetSessionOptionsImpl(int32_t num_threads,
@@ -74,6 +78,38 @@ static Ort::SessionOptions GetSessionOptionsImpl(int32_t num_threads,
                                                             coreml_flags);
 #else
       SHERPA_ONNX_LOGE("CoreML is for Apple only. Fallback to cpu!");
+#endif
+      break;
+    }
+    case Provider::kNNAPI: {
+#if __ANDROID_API__ >= 27
+      SHERPA_ONNX_LOGE("Current API level %d ", (int32_t)__ANDROID_API__);
+
+      // Please see
+      // https://onnxruntime.ai/docs/execution-providers/NNAPI-ExecutionProvider.html#usage
+      // to enable different flags
+      uint32_t nnapi_flags = 0;
+      // nnapi_flags |= NNAPI_FLAG_USE_FP16;
+      // nnapi_flags |= NNAPI_FLAG_CPU_DISABLED;
+      OrtStatus *status = OrtSessionOptionsAppendExecutionProvider_Nnapi(
+          sess_opts, nnapi_flags);
+
+      if (status) {
+        const auto &api = Ort::GetApi();
+        const char *msg = api.GetErrorMessage(status);
+        SHERPA_ONNX_LOGE(
+            "Failed to enable NNAPI: %s. Available providers: %s. Fallback to "
+            "cpu",
+            msg, os.str().c_str());
+        api.ReleaseStatus(status);
+      } else {
+        SHERPA_ONNX_LOGE("Use nnapi");
+      }
+#else
+      SHERPA_ONNX_LOGE(
+          "Android NNAPI requires API level >= 27. Current API level %d "
+          "Fallback to cpu!",
+          (int32_t)__ANDROID_API__);
 #endif
       break;
     }