diff --git a/CMakeLists.txt b/CMakeLists.txt
index edc7989..407a88b 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -30,6 +30,11 @@ if (GGML_CUBLAS)
   set_property(TARGET ggml PROPERTY CUDA_ARCHITECTURES ${CUDA_ARCHITECTURES})
 endif ()
 
+if (GGML_METAL)
+  add_compile_definitions(GGML_USE_METAL)
+  configure_file(third_party/ggml/src/ggml-metal.metal ${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/ggml-metal.metal COPYONLY)
+endif ()
+
 file(GLOB CPP_SOURCES
   ${PROJECT_SOURCE_DIR}/*.h
   ${PROJECT_SOURCE_DIR}/*.cpp)
diff --git a/README.md b/README.md
index 009d09a..7ea7d95 100644
--- a/README.md
+++ b/README.md
@@ -89,6 +89,13 @@ cuBLAS uses NVIDIA GPU to accelerate BLAS. Add the CMake flag `-DGGML_CUBLAS=ON`
 cmake -B build -DGGML_CUBLAS=ON && cmake --build build -j
 ```
 
+**Metal**
+
+MPS (Metal Performance Shaders) allows computation to run on powerful Apple Silicon GPU. Add the CMake flag `-DGGML_METAL=ON` to enable it.
+```sh
+cmake -B build -DGGML_METAL=ON && cmake --build build -j
+```
+
 ## Python Binding
 
 The Python binding provides high-level `chat` and `stream_chat` interface similar to the original Hugging Face Qwen-7B.
diff --git a/qwen.cpp b/qwen.cpp
index 271daa5..910cbf3 100644
--- a/qwen.cpp
+++ b/qwen.cpp
@@ -61,7 +61,21 @@ auto ggml_graph_compute_helper(std::vector<uninitialized_char> &buf, ggml_cgraph
   ggml_graph_compute(graph, &plan);
 }
 
-auto ModelContext::init_device_context() -> void {}
+auto ModelContext::init_device_context() -> void {
+#ifdef GGML_USE_METAL
+  ctx_metal = make_unique_ggml_metal_context(1);
+  const size_t max_size = ggml_get_max_tensor_size(ctx_w.get());
+  void *weight_data = weight_buffer.empty() ? ggml_get_mem_buffer(ctx_w.get()) : (void *)weight_buffer.data();
+  size_t weight_size = weight_buffer.empty() ? ggml_get_mem_size(ctx_w.get()) : weight_buffer.size();
+  QWEN_CHECK(ggml_metal_add_buffer(ctx_metal.get(), "weights", weight_data, weight_size, max_size));
+  QWEN_CHECK(ggml_metal_add_buffer(ctx_metal.get(), "kv", ggml_get_mem_buffer(ctx_kv.get()),
+             ggml_get_mem_size(ctx_kv.get()), 0));
+  void *compute_data = ctx_b ? ggml_get_mem_buffer(ctx_b.get()) : compute_buffer.data();
+  size_t compute_size = ctx_b ? ggml_get_mem_size(ctx_b.get()) : compute_buffer.size();
+  QWEN_CHECK(ggml_metal_add_buffer(ctx_metal.get(), "compute", compute_data, compute_size, 0));
+  QWEN_CHECK(ggml_metal_add_buffer(ctx_metal.get(), "scratch", scratch.data, scratch.size, 0));
+#endif
+}
 
 // ===== streamer =====
 
@@ -482,7 +496,7 @@ auto get_num_physical_cores() -> int {
 }
 
 auto get_default_num_threads() -> int {
-#ifdef GGML_USE_CUBLAS
+#if defined(GGML_USE_CUBLAS) || defined(GGML_USE_METAL)
     return 1;
 #else
   return std::min(get_num_physical_cores(), 16);
@@ -583,7 +597,11 @@ auto QwenForCausalLM::generate_next_token(
   }
 
   ggml_build_forward_expand(&ctx_.gf, lm_logits);
+#ifdef GGML_USE_METAL
+  ggml_metal_graph_compute(ctx_.ctx_metal.get(), &ctx_.gf);
+#else
   ggml_graph_compute_helper(ctx_.work_buffer, &ctx_.gf, n_threads);
+#endif
 
   int vocab_size = lm_logits->ne[0];
   float *next_token_logits = (float *)lm_logits->data;
diff --git a/qwen.h b/qwen.h
index 66bb11c..f8d87cc 100644
--- a/qwen.h
+++ b/qwen.h
@@ -12,6 +12,10 @@
 #include <ggml-cuda.h>
 #endif
 
+#ifdef GGML_USE_METAL
+#include <ggml-metal.h>
+#endif
+
 namespace qwen {
 
 class QwenTokenizer;
@@ -58,6 +62,20 @@ static inline auto make_unique_ggml_context(
   return unique_ggml_context_t(ggml_init({mem_size, mem_buffer, no_alloc}));
 }
 
+#ifdef GGML_USE_METAL
+struct ggml_metal_context_deleter_t {
+  auto operator()(ggml_metal_context *ctx) const noexcept -> void { ggml_metal_free(ctx); }
+};
+
+using unique_ggml_metal_context_t = std::unique_ptr<ggml_metal_context, ggml_metal_context_deleter_t>;
+
+static inline auto make_unique_ggml_metal_context(
+  int n_cb
+) -> unique_ggml_metal_context_t {
+  return unique_ggml_metal_context_t(ggml_metal_init(n_cb));
+}
+#endif
+
 struct uninitialized_char {
   char m;
   uninitialized_char() {}
@@ -70,6 +88,9 @@ struct ModelContext {
   unique_ggml_context_t ctx_w;  // weight
   unique_ggml_context_t ctx_kv; // kv cache
   unique_ggml_context_t ctx_b;  // buffer
+#ifdef GGML_USE_METAL
+  unique_ggml_metal_context_t ctx_metal;
+#endif
   ggml_cgraph gf;
   ggml_scratch scratch;
   std::vector<uninitialized_char> compute_buffer; // BLAS buffer