diff --git a/examples/CMakeLists.txt b/examples/CMakeLists.txt
index ecb70885b..664459997 100644
--- a/examples/CMakeLists.txt
+++ b/examples/CMakeLists.txt
@@ -17,6 +17,7 @@ if(OpenCV_FOUND)
     include_directories(${CMAKE_CURRENT_BINARY_DIR}/../src)
 
     ncnn_add_example(squeezenet)
+    ncnn_add_example(squeezenet_c_api)
     ncnn_add_example(fasterrcnn)
     ncnn_add_example(rfcn)
     ncnn_add_example(yolov2)
diff --git a/examples/squeezenet_c_api.cpp b/examples/squeezenet_c_api.cpp
new file mode 100644
index 000000000..2fe268a05
--- /dev/null
+++ b/examples/squeezenet_c_api.cpp
@@ -0,0 +1,117 @@
+// Tencent is pleased to support the open source community by making ncnn available.
+//
+// Copyright (C) 2020 THL A29 Limited, a Tencent company. All rights reserved.
+//
+// Licensed under the BSD 3-Clause License (the "License"); you may not use this file except
+// in compliance with the License. You may obtain a copy of the License at
+//
+// https://opensource.org/licenses/BSD-3-Clause
+//
+// Unless required by applicable law or agreed to in writing, software distributed
+// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
+// CONDITIONS OF ANY KIND, either express or implied. See the License for the
+// specific language governing permissions and limitations under the License.
+
+#include "c_api.h"
+
+#include <algorithm>
+#include <opencv2/core/core.hpp>
+#include <opencv2/highgui/highgui.hpp>
+#include <stdio.h>
+#include <vector>
+
+static int detect_squeezenet(const cv::Mat& bgr, std::vector<float>& cls_scores)
+{
+    ncnn_net_t squeezenet = ncnn_net_create();
+
+    ncnn_option_t opt = ncnn_option_create();
+    ncnn_option_set_use_vulkan_compute(opt, 1);
+
+    ncnn_net_set_option(squeezenet, opt);
+
+    // the ncnn model https://github.com/nihui/ncnn-assets/tree/master/models
+    ncnn_net_load_param(squeezenet, "squeezenet_v1.1.param");
+    ncnn_net_load_model(squeezenet, "squeezenet_v1.1.bin");
+
+    ncnn_mat_t in = ncnn_mat_from_pixels_resize(bgr.data, NCNN_MAT_PIXEL_BGR, bgr.cols, bgr.rows, bgr.cols * 3, 227, 227);
+
+    const float mean_vals[3] = {104.f, 117.f, 123.f};
+    ncnn_mat_substract_mean_normalize(in, mean_vals, 0);
+
+    ncnn_extractor_t ex = ncnn_extractor_create(squeezenet);
+
+    ncnn_extractor_input(ex, "data", in);
+
+    ncnn_mat_t out;
+    ncnn_extractor_extract(ex, "prob", &out);
+
+    const int out_w = ncnn_mat_get_w(out);
+    const float* out_data = (const float*)ncnn_mat_get_data(out);
+
+    cls_scores.resize(out_w);
+    for (int j = 0; j < out_w; j++)
+    {
+        cls_scores[j] = out_data[j];
+    }
+
+    ncnn_mat_destroy(in);
+    ncnn_mat_destroy(out);
+
+    ncnn_extractor_destroy(ex);
+
+    ncnn_option_destroy(opt);
+
+    ncnn_net_destroy(squeezenet);
+
+    return 0;
+}
+
+static int print_topk(const std::vector<float>& cls_scores, int topk)
+{
+    // partial sort topk with index
+    int size = cls_scores.size();
+    std::vector<std::pair<float, int> > vec;
+    vec.resize(size);
+    for (int i = 0; i < size; i++)
+    {
+        vec[i] = std::make_pair(cls_scores[i], i);
+    }
+
+    std::partial_sort(vec.begin(), vec.begin() + topk, vec.end(),
+                      std::greater<std::pair<float, int> >());
+
+    // print topk and score
+    for (int i = 0; i < topk; i++)
+    {
+        float score = vec[i].first;
+        int index = vec[i].second;
+        fprintf(stderr, "%d = %f\n", index, score);
+    }
+
+    return 0;
+}
+
+int main(int argc, char** argv)
+{
+    if (argc != 2)
+    {
+        fprintf(stderr, "Usage: %s [imagepath]\n", argv[0]);
+        return -1;
+    }
+
+    const char* imagepath = argv[1];
+
+    cv::Mat m = cv::imread(imagepath, 1);
+    if (m.empty())
+    {
+        fprintf(stderr, "cv::imread %s failed\n", imagepath);
+        return -1;
+    }
+
+    std::vector<float> cls_scores;
+    detect_squeezenet(m, cls_scores);
+
+    print_topk(cls_scores, 3);
+
+    return 0;
+}
diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index 4b69ab090..bea9fe38d 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -23,6 +23,7 @@ set(ncnn_SRCS
     allocator.cpp
     benchmark.cpp
     blob.cpp
+    c_api.cpp
     command.cpp
     cpu.cpp
     datareader.cpp
@@ -425,6 +426,7 @@ if(NCNN_INSTALL_SDK)
     install(FILES
         allocator.h
         blob.h
+        c_api.h
         command.h
         cpu.h
         datareader.h
diff --git a/src/c_api.cpp b/src/c_api.cpp
new file mode 100644
index 000000000..34224b6aa
--- /dev/null
+++ b/src/c_api.cpp
@@ -0,0 +1,372 @@
+/* Tencent is pleased to support the open source community by making ncnn available.
+ *
+ * Copyright (C) 2020 THL A29 Limited, a Tencent company. All rights reserved.
+ *
+ * Licensed under the BSD 3-Clause License (the "License"); you may not use this file except
+ * in compliance with the License. You may obtain a copy of the License at
+ *
+ * https://opensource.org/licenses/BSD-3-Clause
+ *
+ * Unless required by applicable law or agreed to in writing, software distributed
+ * under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
+ * CONDITIONS OF ANY KIND, either express or implied. See the License for the
+ * specific language governing permissions and limitations under the License.
+ */
+
+#include "c_api.h"
+
+#include "mat.h"
+#include "option.h"
+#include "net.h"
+
+using ncnn::Blob;
+using ncnn::Extractor;
+using ncnn::Layer;
+using ncnn::Mat;
+using ncnn::Net;
+using ncnn::Option;
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* mat api */
+ncnn_mat_t ncnn_mat_create()
+{
+    return (ncnn_mat_t)(new Mat());
+}
+
+ncnn_mat_t ncnn_mat_create_1d(int w)
+{
+    return (ncnn_mat_t)(new Mat(w));
+}
+
+ncnn_mat_t ncnn_mat_create_2d(int w, int h)
+{
+    return (ncnn_mat_t)(new Mat(w, h));
+}
+
+ncnn_mat_t ncnn_mat_create_3d(int w, int h, int c)
+{
+    return (ncnn_mat_t)(new Mat(w, h, c));
+}
+
+ncnn_mat_t ncnn_mat_create_1d_packed(int w, size_t elemsize, int elempack)
+{
+    return (ncnn_mat_t)(new Mat(w, elemsize, elempack));
+}
+
+ncnn_mat_t ncnn_mat_create_2d_packed(int w, int h, size_t elemsize, int elempack)
+{
+    return (ncnn_mat_t)(new Mat(w, h, elemsize, elempack));
+}
+
+ncnn_mat_t ncnn_mat_create_3d_packed(int w, int h, int c, size_t elemsize, int elempack)
+{
+    return (ncnn_mat_t)(new Mat(w, h, c, elemsize, elempack));
+}
+
+void ncnn_mat_destroy(ncnn_mat_t mat)
+{
+    delete (Mat*)mat;
+}
+
+int ncnn_mat_get_dims(ncnn_mat_t mat)
+{
+    return ((Mat*)mat)->dims;
+}
+
+int ncnn_mat_get_w(ncnn_mat_t mat)
+{
+    return ((Mat*)mat)->w;
+}
+
+int ncnn_mat_get_h(ncnn_mat_t mat)
+{
+    return ((Mat*)mat)->h;
+}
+
+int ncnn_mat_get_c(ncnn_mat_t mat)
+{
+    return ((Mat*)mat)->c;
+}
+
+size_t ncnn_mat_get_elemsize(ncnn_mat_t mat)
+{
+    return ((Mat*)mat)->elemsize;
+}
+
+int ncnn_mat_get_elempack(ncnn_mat_t mat)
+{
+    return ((Mat*)mat)->elempack;
+}
+
+size_t ncnn_mat_get_cstep(ncnn_mat_t mat)
+{
+    return ((Mat*)mat)->cstep;
+}
+
+void* ncnn_mat_get_data(ncnn_mat_t mat)
+{
+    return ((Mat*)mat)->data;
+}
+
+/* mat pixel api */
+ncnn_mat_t ncnn_mat_from_pixels(const unsigned char* pixels, int type, int w, int h, int stride)
+{
+    return (ncnn_mat_t)(new Mat(Mat::from_pixels(pixels, type, w, h, stride)));
+}
+
+ncnn_mat_t ncnn_mat_from_pixels_resize(const unsigned char* pixels, int type, int w, int h, int stride, int target_width, int target_height)
+{
+    return (ncnn_mat_t)(new Mat(Mat::from_pixels_resize(pixels, type, w, h, stride, target_width, target_height)));
+}
+
+void ncnn_mat_to_pixels(ncnn_mat_t mat, unsigned char* pixels, int type, int stride)
+{
+    ((Mat*)mat)->to_pixels(pixels, type, stride);
+}
+
+void ncnn_mat_to_pixels_resize(ncnn_mat_t mat, unsigned char* pixels, int type, int target_width, int target_height, int target_stride)
+{
+    ((Mat*)mat)->to_pixels_resize(pixels, type, target_width, target_height, target_stride);
+}
+
+void ncnn_mat_substract_mean_normalize(ncnn_mat_t mat, const float* mean_vals, const float* norm_vals)
+{
+    ((Mat*)mat)->substract_mean_normalize(mean_vals, norm_vals);
+}
+
+/* option api */
+ncnn_option_t ncnn_option_create()
+{
+    return (ncnn_option_t)(new Option());
+}
+
+void ncnn_option_destroy(ncnn_option_t opt)
+{
+    delete (Option*)opt;
+}
+
+int ncnn_option_get_num_threads(ncnn_option_t opt)
+{
+    return ((Option*)opt)->num_threads;
+}
+
+void ncnn_option_set_num_threads(ncnn_option_t opt, int num_threads)
+{
+    ((Option*)opt)->num_threads = num_threads;
+}
+
+int ncnn_option_get_use_vulkan_compute(ncnn_option_t opt)
+{
+#if NCNN_VULKAN
+    return ((Option*)opt)->use_vulkan_compute;
+#else
+    return 0;
+#endif
+}
+
+void ncnn_option_set_use_vulkan_compute(ncnn_option_t opt, int use_vulkan_compute)
+{
+#if NCNN_VULKAN
+    ((Option*)opt)->use_vulkan_compute = use_vulkan_compute;
+#else
+    (void)opt;
+    (void)use_vulkan_compute;
+#endif
+}
+
+/* blob api */
+const char* ncnn_blob_get_name(ncnn_blob_t blob)
+{
+#if NCNN_STRING
+    return ((Blob*)blob)->name.c_str();
+#else
+    return "";
+#endif
+}
+
+int ncnn_blob_get_producer(ncnn_blob_t blob)
+{
+    return ((Blob*)blob)->producer;
+}
+
+int ncnn_blob_get_consumer_count(ncnn_blob_t blob)
+{
+    return (int)((Blob*)blob)->consumers.size();
+}
+
+int ncnn_blob_get_consumer(ncnn_blob_t blob, int i)
+{
+    return ((Blob*)blob)->consumers[i];
+}
+
+void ncnn_blob_get_shape(ncnn_blob_t blob, int* dims, int* w, int* h, int* c)
+{
+    const Mat& shape = ((Blob*)blob)->shape;
+    *dims = shape.dims;
+    *w = shape.w;
+    *h = shape.h;
+    *c = shape.c;
+}
+
+/* layer api */
+const char* ncnn_layer_get_name(ncnn_layer_t layer)
+{
+#if NCNN_STRING
+    return ((Layer*)layer)->name.c_str();
+#else
+    return "";
+#endif
+}
+
+int ncnn_layer_get_typeindex(ncnn_layer_t layer)
+{
+    return ((Layer*)layer)->typeindex;
+}
+
+const char* ncnn_layer_get_type(ncnn_layer_t layer)
+{
+#if NCNN_STRING
+    return ((Layer*)layer)->type.c_str();
+#else
+    return "";
+#endif
+}
+
+int ncnn_layer_get_bottom_count(ncnn_layer_t layer)
+{
+    return (int)((Layer*)layer)->bottoms.size();
+}
+
+int ncnn_layer_get_bottom(ncnn_layer_t layer, int i)
+{
+    return ((Layer*)layer)->bottoms[i];
+}
+
+int ncnn_layer_get_top_count(ncnn_layer_t layer)
+{
+    return (int)((Layer*)layer)->tops.size();
+}
+
+int ncnn_layer_get_top(ncnn_layer_t layer, int i)
+{
+    return ((Layer*)layer)->tops[i];
+}
+
+void ncnn_blob_get_bottom_shape(ncnn_layer_t layer, int i, int* dims, int* w, int* h, int* c)
+{
+    const Mat& shape = ((Layer*)layer)->bottom_shapes[i];
+    *dims = shape.dims;
+    *w = shape.w;
+    *h = shape.h;
+    *c = shape.c;
+}
+
+void ncnn_blob_get_top_shape(ncnn_layer_t layer, int i, int* dims, int* w, int* h, int* c)
+{
+    const Mat& shape = ((Layer*)layer)->top_shapes[i];
+    *dims = shape.dims;
+    *w = shape.w;
+    *h = shape.h;
+    *c = shape.c;
+}
+
+/* net api */
+ncnn_net_t ncnn_net_create()
+{
+    return (ncnn_net_t)(new Net);
+}
+
+void ncnn_net_destroy(ncnn_net_t net)
+{
+    delete (Net*)net;
+}
+
+void ncnn_net_set_option(ncnn_net_t net, ncnn_option_t opt)
+{
+    ((Net*)net)->opt = *((Option*)opt);
+}
+
+int ncnn_net_load_param(ncnn_net_t net, const char* path)
+{
+#if NCNN_STDIO && NCNN_STRING
+    return ((Net*)net)->load_param(path);
+#else
+    return -1;
+#endif
+}
+
+int ncnn_net_load_model(ncnn_net_t net, const char* path)
+{
+#if NCNN_STDIO && NCNN_STRING
+    return ((Net*)net)->load_model(path);
+#else
+    return -1;
+#endif
+}
+
+int ncnn_net_get_layer_count(ncnn_net_t net)
+{
+    return (int)((Net*)net)->layers.size();
+}
+
+ncnn_layer_t ncnn_net_get_layer(ncnn_net_t net, int i)
+{
+    return (ncnn_layer_t)((Net*)net)->layers[i];
+}
+
+int ncnn_net_get_blob_count(ncnn_net_t net)
+{
+    return (int)((Net*)net)->blobs.size();
+}
+
+ncnn_blob_t ncnn_net_get_blob(ncnn_net_t net, int i)
+{
+    return (ncnn_blob_t) & ((Net*)net)->blobs[i];
+}
+
+/* extractor api */
+ncnn_extractor_t ncnn_extractor_create(ncnn_net_t net)
+{
+    return (ncnn_extractor_t)(new Extractor(((Net*)net)->create_extractor()));
+}
+
+void ncnn_extractor_destroy(ncnn_extractor_t ex)
+{
+    delete (Extractor*)ex;
+}
+
+void ncnn_extractor_set_option(ncnn_extractor_t ex, ncnn_option_t opt)
+{
+    ((Extractor*)ex)->set_num_threads(((Option*)opt)->num_threads);
+#if NCNN_VULKAN
+    ((Extractor*)ex)->set_vulkan_compute(((Option*)opt)->use_vulkan_compute);
+#endif
+}
+
+int ncnn_extractor_input(ncnn_extractor_t ex, const char* name, ncnn_mat_t mat)
+{
+#if NCNN_STRING
+    return ((Extractor*)ex)->input(name, *((Mat*)mat));
+#else
+    return -1;
+#endif
+}
+
+int ncnn_extractor_extract(ncnn_extractor_t ex, const char* name, ncnn_mat_t* mat)
+{
+#if NCNN_STRING
+    Mat mat0;
+    int ret = ((Extractor*)ex)->extract(name, mat0);
+    *mat = (ncnn_mat_t)(new Mat(mat0));
+    return ret;
+#else
+    return -1;
+#endif
+}
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif
diff --git a/src/c_api.h b/src/c_api.h
new file mode 100644
index 000000000..1ed4b7f05
--- /dev/null
+++ b/src/c_api.h
@@ -0,0 +1,130 @@
+/* Tencent is pleased to support the open source community by making ncnn available.
+ *
+ * Copyright (C) 2020 THL A29 Limited, a Tencent company. All rights reserved.
+ *
+ * Licensed under the BSD 3-Clause License (the "License"); you may not use this file except
+ * in compliance with the License. You may obtain a copy of the License at
+ *
+ * https://opensource.org/licenses/BSD-3-Clause
+ *
+ * Unless required by applicable law or agreed to in writing, software distributed
+ * under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
+ * CONDITIONS OF ANY KIND, either express or implied. See the License for the
+ * specific language governing permissions and limitations under the License.
+ */
+
+#ifndef NCNN_C_API_H
+#define NCNN_C_API_H
+
+#include <stddef.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/* mat api */
+typedef struct __ncnn_mat_t* ncnn_mat_t;
+
+ncnn_mat_t ncnn_mat_create();
+ncnn_mat_t ncnn_mat_create_1d(int w);
+ncnn_mat_t ncnn_mat_create_2d(int w, int h);
+ncnn_mat_t ncnn_mat_create_3d(int w, int h, int c);
+ncnn_mat_t ncnn_mat_create_1d_packed(int w, size_t elemsize, int elempack);
+ncnn_mat_t ncnn_mat_create_2d_packed(int w, int h, size_t elemsize, int elempack);
+ncnn_mat_t ncnn_mat_create_3d_packed(int w, int h, int c, size_t elemsize, int elempack);
+void ncnn_mat_destroy(ncnn_mat_t mat);
+
+int ncnn_mat_get_dims(ncnn_mat_t mat);
+int ncnn_mat_get_w(ncnn_mat_t mat);
+int ncnn_mat_get_h(ncnn_mat_t mat);
+int ncnn_mat_get_c(ncnn_mat_t mat);
+size_t ncnn_mat_get_elemsize(ncnn_mat_t mat);
+int ncnn_mat_get_elempack(ncnn_mat_t mat);
+size_t ncnn_mat_get_cstep(ncnn_mat_t mat);
+void* ncnn_mat_get_data(ncnn_mat_t mat);
+
+/* mat pixel api */
+#define NCNN_MAT_PIXEL_RGB       1
+#define NCNN_MAT_PIXEL_BGR       2
+#define NCNN_MAT_PIXEL_GRAY      3
+#define NCNN_MAT_PIXEL_RGBA      4
+#define NCNN_MAT_PIXEL_BGRA      5
+#define NCNN_MAT_PIXEL_X2Y(X, Y) (X | (Y << 16))
+ncnn_mat_t ncnn_mat_from_pixels(const unsigned char* pixels, int type, int w, int h, int stride);
+ncnn_mat_t ncnn_mat_from_pixels_resize(const unsigned char* pixels, int type, int w, int h, int stride, int target_width, int target_height);
+void ncnn_mat_to_pixels(ncnn_mat_t mat, unsigned char* pixels, int type, int stride);
+void ncnn_mat_to_pixels_resize(ncnn_mat_t mat, unsigned char* pixels, int type, int target_width, int target_height, int target_stride);
+
+void ncnn_mat_substract_mean_normalize(ncnn_mat_t mat, const float* mean_vals, const float* norm_vals);
+
+/* option api */
+typedef struct __ncnn_option_t* ncnn_option_t;
+
+ncnn_option_t ncnn_option_create();
+void ncnn_option_destroy(ncnn_option_t opt);
+
+int ncnn_option_get_num_threads(ncnn_option_t opt);
+void ncnn_option_set_num_threads(ncnn_option_t opt, int num_threads);
+
+int ncnn_option_get_use_vulkan_compute(ncnn_option_t opt);
+void ncnn_option_set_use_vulkan_compute(ncnn_option_t opt, int use_vulkan_compute);
+
+/* blob api */
+typedef struct __ncnn_blob_t* ncnn_blob_t;
+
+const char* ncnn_blob_get_name(ncnn_blob_t blob);
+
+int ncnn_blob_get_producer(ncnn_blob_t blob);
+int ncnn_blob_get_consumer_count(ncnn_blob_t blob);
+int ncnn_blob_get_consumer(ncnn_blob_t blob, int i);
+
+void ncnn_blob_get_shape(ncnn_blob_t blob, int* dims, int* w, int* h, int* c);
+
+/* layer api */
+typedef struct __ncnn_layer_t* ncnn_layer_t;
+
+const char* ncnn_layer_get_name(ncnn_layer_t layer);
+
+int ncnn_layer_get_typeindex(ncnn_layer_t layer);
+const char* ncnn_layer_get_type(ncnn_layer_t layer);
+
+int ncnn_layer_get_bottom_count(ncnn_layer_t layer);
+int ncnn_layer_get_bottom(ncnn_layer_t layer, int i);
+int ncnn_layer_get_top_count(ncnn_layer_t layer);
+int ncnn_layer_get_top(ncnn_layer_t layer, int i);
+
+void ncnn_blob_get_bottom_shape(ncnn_layer_t layer, int i, int* dims, int* w, int* h, int* c);
+void ncnn_blob_get_top_shape(ncnn_layer_t layer, int i, int* dims, int* w, int* h, int* c);
+
+/* net api */
+typedef struct __ncnn_net_t* ncnn_net_t;
+
+ncnn_net_t ncnn_net_create();
+void ncnn_net_destroy(ncnn_net_t net);
+
+void ncnn_net_set_option(ncnn_net_t net, ncnn_option_t opt);
+
+int ncnn_net_load_param(ncnn_net_t net, const char* path);
+int ncnn_net_load_model(ncnn_net_t net, const char* path);
+
+int ncnn_net_get_layer_count(ncnn_net_t net);
+ncnn_layer_t ncnn_net_get_layer(ncnn_net_t net, int i);
+int ncnn_net_get_blob_count(ncnn_net_t net);
+ncnn_blob_t ncnn_net_get_blob(ncnn_net_t net, int i);
+
+/* extractor api */
+typedef struct __ncnn_extractor_t* ncnn_extractor_t;
+
+ncnn_extractor_t ncnn_extractor_create(ncnn_net_t net);
+void ncnn_extractor_destroy(ncnn_extractor_t ex);
+
+void ncnn_extractor_set_option(ncnn_extractor_t ex, ncnn_option_t opt);
+
+int ncnn_extractor_input(ncnn_extractor_t ex, const char* name, ncnn_mat_t mat);
+int ncnn_extractor_extract(ncnn_extractor_t ex, const char* name, ncnn_mat_t* mat);
+
+#ifdef __cplusplus
+} /* extern "C" */
+#endif
+
+#endif // NCNN_C_API_H
diff --git a/src/net.h b/src/net.h
index 93b0a3bd9..c450c3b92 100644
--- a/src/net.h
+++ b/src/net.h
@@ -123,6 +123,10 @@ public:
     // construct an Extractor from network
     Extractor create_extractor() const;
 
+public:
+    std::vector<Blob> blobs;
+    std::vector<Layer*> layers;
+
 protected:
     // parse the structure of network
     // fuse int8 op dequantize and quantize by requantize
@@ -154,9 +158,6 @@ protected:
 #endif // NCNN_VULKAN
 
 protected:
-    std::vector<Blob> blobs;
-    std::vector<Layer*> layers;
-
     std::vector<layer_registry_entry> custom_layer_registry;
 
 #if NCNN_VULKAN