diff --git a/src/renderer/wg_engine/meson.build b/src/renderer/wg_engine/meson.build
index d086e646..5405b559 100644
--- a/src/renderer/wg_engine/meson.build
+++ b/src/renderer/wg_engine/meson.build
@@ -3,6 +3,7 @@ source_file = [
     'tvgWgPipelineEmpty.cpp',
     'tvgWgPipelineLinear.cpp',
     'tvgWgPipelineSolid.cpp',
+    'tvgWgPipelineRadial.cpp',
     'tvgWgRenderData.cpp',
     'tvgWgRenderer.cpp',
     'tvgWgShaderSrc.cpp'
diff --git a/src/renderer/wg_engine/tvgWgPipelineRadial.cpp b/src/renderer/wg_engine/tvgWgPipelineRadial.cpp
new file mode 100644
index 00000000..64b15231
--- /dev/null
+++ b/src/renderer/wg_engine/tvgWgPipelineRadial.cpp
@@ -0,0 +1,303 @@
+/*
+ * Copyright (c) 2023 the ThorVG project. All rights reserved.
+
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include "tvgWgPipelineRadial.h"
+#include "tvgWgShaderSrc.h"
+
+//************************************************************************
+// WgPipelineBindGroupRadial
+//************************************************************************
+
+void WgPipelineDataRadial::updateGradient(RadialGradient* radialGradient) {
+    const Fill::ColorStop* stops = nullptr;
+    auto stopCnt = radialGradient->colorStops(&stops);
+
+    uGradientInfo.nStops[0] = stopCnt;
+    uGradientInfo.nStops[1] = 0.5f;
+
+    for (uint32_t i = 0; i < stopCnt; ++i) {
+        uGradientInfo.stopPoints[i] = stops[i].offset;
+        uGradientInfo.stopColors[i * 4 + 0] = stops[i].r / 255.f;
+        uGradientInfo.stopColors[i * 4 + 1] = stops[i].g / 255.f;
+        uGradientInfo.stopColors[i * 4 + 2] = stops[i].b / 255.f;
+        uGradientInfo.stopColors[i * 4 + 3] = stops[i].a / 255.f;
+    }
+
+    radialGradient->radial(
+        &uGradientInfo.centerPos[0],
+        &uGradientInfo.centerPos[1],
+        &uGradientInfo.radius[0]);
+}
+
+//************************************************************************
+// WgPipelineBindGroupRadial
+//************************************************************************
+
+void WgPipelineBindGroupRadial::initialize(WGPUDevice device, WgPipelineRadial& pipelinePipelineRadial) {
+        // buffer uniform uMatrix
+    WGPUBufferDescriptor bufferUniformDesc_uMatrix{};
+    bufferUniformDesc_uMatrix.nextInChain = nullptr;
+    bufferUniformDesc_uMatrix.label = "Buffer uniform pipeline radial uMatrix";
+    bufferUniformDesc_uMatrix.usage = WGPUBufferUsage_CopyDst | WGPUBufferUsage_Uniform;
+    bufferUniformDesc_uMatrix.size = sizeof(WgPipelineMatrix);
+    bufferUniformDesc_uMatrix.mappedAtCreation = false;
+    uBufferMatrix = wgpuDeviceCreateBuffer(device, &bufferUniformDesc_uMatrix);
+    assert(uBufferMatrix);
+    // buffer uniform uColorInfo
+    WGPUBufferDescriptor bufferUniformDesc_uGradientInfo{};
+    bufferUniformDesc_uGradientInfo.nextInChain = nullptr;
+    bufferUniformDesc_uGradientInfo.label = "Buffer uniform pipeline radial uGradientInfo";
+    bufferUniformDesc_uGradientInfo.usage = WGPUBufferUsage_CopyDst | WGPUBufferUsage_Uniform;
+    bufferUniformDesc_uGradientInfo.size = sizeof(WgPipelineRadialGradientInfo);
+    bufferUniformDesc_uGradientInfo.mappedAtCreation = false;
+    uBufferGradientInfo = wgpuDeviceCreateBuffer(device, &bufferUniformDesc_uGradientInfo);
+    assert(uBufferGradientInfo);
+
+    // bind group entry @binding(0) uMatrix
+    WGPUBindGroupEntry bindGroupEntry_uMatrix{};
+    bindGroupEntry_uMatrix.nextInChain = nullptr;
+    bindGroupEntry_uMatrix.binding = 0;
+    bindGroupEntry_uMatrix.buffer = uBufferMatrix;
+    bindGroupEntry_uMatrix.offset = 0;
+    bindGroupEntry_uMatrix.size = sizeof(WgPipelineMatrix);
+    bindGroupEntry_uMatrix.sampler = nullptr;
+    bindGroupEntry_uMatrix.textureView = nullptr;
+    // bind group entry @binding(1) uGradientInfo
+    WGPUBindGroupEntry bindGroupEntry_uGradientInfo{};
+    bindGroupEntry_uGradientInfo.nextInChain = nullptr;
+    bindGroupEntry_uGradientInfo.binding = 1;
+    bindGroupEntry_uGradientInfo.buffer = uBufferGradientInfo;
+    bindGroupEntry_uGradientInfo.offset = 0;
+    bindGroupEntry_uGradientInfo.size = sizeof(WgPipelineRadialGradientInfo);
+    bindGroupEntry_uGradientInfo.sampler = nullptr;
+    bindGroupEntry_uGradientInfo.textureView = nullptr;
+    // bind group entries
+    WGPUBindGroupEntry bindGroupEntries[] {
+        bindGroupEntry_uMatrix,      // @binding(0) uMatrix
+        bindGroupEntry_uGradientInfo // @binding(1) uGradientInfo
+    };
+    // bind group descriptor
+    WGPUBindGroupDescriptor bindGroupDescPipeline{};
+    bindGroupDescPipeline.nextInChain = nullptr;
+    bindGroupDescPipeline.label = "The binding group pipeline radial";
+    bindGroupDescPipeline.layout = pipelinePipelineRadial.mBindGroupLayout;
+    bindGroupDescPipeline.entryCount = 2;
+    bindGroupDescPipeline.entries = bindGroupEntries;
+    mBindGroup = wgpuDeviceCreateBindGroup(device, &bindGroupDescPipeline);
+    assert(mBindGroup);
+}
+
+void WgPipelineBindGroupRadial::release() {
+    if (uBufferGradientInfo) {
+        wgpuBufferDestroy(uBufferGradientInfo);
+        wgpuBufferRelease(uBufferGradientInfo);
+        uBufferGradientInfo = nullptr;
+    }
+    if (uBufferMatrix) {
+        wgpuBufferDestroy(uBufferMatrix);
+        wgpuBufferRelease(uBufferMatrix);
+        uBufferMatrix = nullptr;
+    }
+    if (mBindGroup) {
+        wgpuBindGroupRelease(mBindGroup);
+        mBindGroup = nullptr;
+    }
+}
+
+void WgPipelineBindGroupRadial::update(WGPUQueue queue, WgPipelineDataRadial& pipelineDataRadial) {
+    wgpuQueueWriteBuffer(queue, uBufferMatrix, 0, &pipelineDataRadial.uMatrix, sizeof(pipelineDataRadial.uMatrix));
+    wgpuQueueWriteBuffer(queue, uBufferGradientInfo, 0, &pipelineDataRadial.uGradientInfo, sizeof(pipelineDataRadial.uGradientInfo));
+}
+
+//************************************************************************
+// WgPipelineRadial
+//************************************************************************
+
+void WgPipelineRadial::initialize(WGPUDevice device) {
+    // bind group layout group 0
+    // bind group layout descriptor @group(0) @binding(0) uMatrix
+    WGPUBindGroupLayoutEntry bindGroupLayoutEntry_uMatrix{};
+    bindGroupLayoutEntry_uMatrix.nextInChain = nullptr;
+    bindGroupLayoutEntry_uMatrix.binding = 0;
+    bindGroupLayoutEntry_uMatrix.visibility = WGPUShaderStage_Vertex | WGPUShaderStage_Fragment;
+    bindGroupLayoutEntry_uMatrix.buffer.nextInChain = nullptr;
+    bindGroupLayoutEntry_uMatrix.buffer.type = WGPUBufferBindingType_Uniform;
+    bindGroupLayoutEntry_uMatrix.buffer.hasDynamicOffset = false;
+    bindGroupLayoutEntry_uMatrix.buffer.minBindingSize = 0;
+    // bind group layout descriptor @group(0) @binding(1) uColorInfo
+    WGPUBindGroupLayoutEntry bindGroupLayoutEntry_uColorInfo{};
+    bindGroupLayoutEntry_uColorInfo.nextInChain = nullptr;
+    bindGroupLayoutEntry_uColorInfo.binding = 1;
+    bindGroupLayoutEntry_uColorInfo.visibility = WGPUShaderStage_Vertex | WGPUShaderStage_Fragment;
+    bindGroupLayoutEntry_uColorInfo.buffer.nextInChain = nullptr;
+    bindGroupLayoutEntry_uColorInfo.buffer.type = WGPUBufferBindingType_Uniform;
+    bindGroupLayoutEntry_uColorInfo.buffer.hasDynamicOffset = false;
+    bindGroupLayoutEntry_uColorInfo.buffer.minBindingSize = 0;
+    // bind group layout entries @group(0)
+    WGPUBindGroupLayoutEntry bindGroupLayoutEntries[] {
+        bindGroupLayoutEntry_uMatrix,
+        bindGroupLayoutEntry_uColorInfo
+    };
+    // bind group layout descriptor scene @group(0)
+    WGPUBindGroupLayoutDescriptor bindGroupLayoutDesc{};
+    bindGroupLayoutDesc.nextInChain = nullptr;
+    bindGroupLayoutDesc.label = "Bind group layout pipeline radial";
+    bindGroupLayoutDesc.entryCount = 2;
+    bindGroupLayoutDesc.entries = bindGroupLayoutEntries; // @binding
+    mBindGroupLayout = wgpuDeviceCreateBindGroupLayout(device, &bindGroupLayoutDesc);
+    assert(mBindGroupLayout);
+
+    // pipeline layout
+
+    // bind group layout descriptors
+    WGPUBindGroupLayout mBindGroupLayouts[] {
+        mBindGroupLayout
+    };
+    // pipeline layout descriptor
+    WGPUPipelineLayoutDescriptor pipelineLayoutDesc{};
+    pipelineLayoutDesc.nextInChain = nullptr;
+    pipelineLayoutDesc.label = "Pipeline pipeline layout radial";
+    pipelineLayoutDesc.bindGroupLayoutCount = 1;
+    pipelineLayoutDesc.bindGroupLayouts = mBindGroupLayouts;
+    mPipelineLayout = wgpuDeviceCreatePipelineLayout(device, &pipelineLayoutDesc);
+    assert(mPipelineLayout);
+
+    // depth stencil state
+    WGPUDepthStencilState depthStencilState{};
+    depthStencilState.nextInChain = nullptr;
+    depthStencilState.format = WGPUTextureFormat_Stencil8;
+    depthStencilState.depthWriteEnabled = false;
+    depthStencilState.depthCompare = WGPUCompareFunction_Always;
+    // depthStencilState.stencilFront
+    depthStencilState.stencilFront.compare = WGPUCompareFunction_NotEqual;
+    depthStencilState.stencilFront.failOp = WGPUStencilOperation_Zero;
+    depthStencilState.stencilFront.depthFailOp = WGPUStencilOperation_Zero;
+    depthStencilState.stencilFront.passOp = WGPUStencilOperation_Zero;
+    // depthStencilState.stencilBack
+    depthStencilState.stencilBack.compare = WGPUCompareFunction_NotEqual;
+    depthStencilState.stencilBack.failOp = WGPUStencilOperation_Zero;
+    depthStencilState.stencilBack.depthFailOp = WGPUStencilOperation_Zero;
+    depthStencilState.stencilBack.passOp = WGPUStencilOperation_Zero;
+    // stencil mask
+    depthStencilState.stencilReadMask = 0xFFFFFFFF;
+    depthStencilState.stencilWriteMask = 0xFFFFFFFF;
+    // depth bias
+    depthStencilState.depthBias = 0;
+    depthStencilState.depthBiasSlopeScale = 0.0f;
+    depthStencilState.depthBiasClamp = 0.0f;
+
+    // shader module wgsl descriptor
+    WGPUShaderModuleWGSLDescriptor shaderModuleWGSLDesc{};
+	shaderModuleWGSLDesc.chain.next = nullptr;
+	shaderModuleWGSLDesc.chain.sType = WGPUSType_ShaderModuleWGSLDescriptor;
+    shaderModuleWGSLDesc.code = cShaderSource_PipelineRadial;
+    // shader module descriptor
+    WGPUShaderModuleDescriptor shaderModuleDesc{};
+    shaderModuleDesc.nextInChain = &shaderModuleWGSLDesc.chain;
+    shaderModuleDesc.label = "The shader module pipeline radial";
+    shaderModuleDesc.hintCount = 0;
+    shaderModuleDesc.hints = nullptr;
+    mShaderModule = wgpuDeviceCreateShaderModule(device, &shaderModuleDesc);
+    assert(mShaderModule);
+
+    // vertex attributes
+    WGPUVertexAttribute vertexAttributes[] = {
+        { WGPUVertexFormat_Float32x3, sizeof(float) * 0, 0 }, // position
+    };
+    // vertex buffer layout
+    WGPUVertexBufferLayout vertexBufferLayout{};
+    vertexBufferLayout.arrayStride = sizeof(float) * 3; // position
+    vertexBufferLayout.stepMode = WGPUVertexStepMode_Vertex;
+    vertexBufferLayout.attributeCount = 1; // position
+    vertexBufferLayout.attributes = vertexAttributes;
+    
+    // blend state
+    WGPUBlendState blendState{};
+    // blendState.color
+    blendState.color.operation = WGPUBlendOperation_Add;
+    blendState.color.srcFactor = WGPUBlendFactor_SrcAlpha;
+    blendState.color.dstFactor = WGPUBlendFactor_OneMinusSrcAlpha;
+    // blendState.alpha
+    blendState.alpha.operation = WGPUBlendOperation_Add;
+    blendState.alpha.srcFactor = WGPUBlendFactor_Zero;
+    blendState.alpha.dstFactor = WGPUBlendFactor_One;
+
+    // color target state (WGPUTextureFormat_BGRA8UnormSrgb)
+    WGPUColorTargetState colorTargetState0{};
+    colorTargetState0.nextInChain = nullptr;
+    colorTargetState0.format = WGPUTextureFormat_BGRA8Unorm;
+    //colorTargetState0.format = WGPUTextureFormat_BGRA8UnormSrgb;
+    colorTargetState0.blend = &blendState;
+    colorTargetState0.writeMask = WGPUColorWriteMask_All;
+    // color target states
+    WGPUColorTargetState colorTargetStates[] = {
+        colorTargetState0
+    };
+    // fragmanet state
+    WGPUFragmentState fragmentState{};
+    fragmentState.nextInChain = nullptr;
+    fragmentState.module = mShaderModule;
+    fragmentState.entryPoint = "fs_main";
+    fragmentState.constantCount = 0;
+    fragmentState.constants = nullptr;
+    fragmentState.targetCount = 1;
+    fragmentState.targets = colorTargetStates; // render target index
+
+    // render pipeline descriptor
+    WGPURenderPipelineDescriptor renderPipelineDesc{};
+    renderPipelineDesc.nextInChain = nullptr;
+    renderPipelineDesc.label = "Render pipeline pipeline radial";
+    // renderPipelineDesc.layout
+    renderPipelineDesc.layout = mPipelineLayout;
+    // renderPipelineDesc.vertex
+    renderPipelineDesc.vertex.nextInChain = nullptr;
+    renderPipelineDesc.vertex.module = mShaderModule;
+    renderPipelineDesc.vertex.entryPoint = "vs_main";
+    renderPipelineDesc.vertex.constantCount = 0;
+    renderPipelineDesc.vertex.constants = nullptr;
+    renderPipelineDesc.vertex.bufferCount = 1;
+    renderPipelineDesc.vertex.buffers = &vertexBufferLayout; // buffer index
+    // renderPipelineDesc.primitive
+    renderPipelineDesc.primitive.nextInChain = nullptr;
+    renderPipelineDesc.primitive.topology = WGPUPrimitiveTopology_TriangleList;
+    renderPipelineDesc.primitive.stripIndexFormat = WGPUIndexFormat_Undefined;
+    renderPipelineDesc.primitive.frontFace = WGPUFrontFace_CCW;
+    renderPipelineDesc.primitive.cullMode = WGPUCullMode_None;
+    // renderPipelineDesc.depthStencil
+    renderPipelineDesc.depthStencil = &depthStencilState;
+    // renderPipelineDesc.multisample
+    renderPipelineDesc.multisample.nextInChain = nullptr;
+    renderPipelineDesc.multisample.count = 1;
+    renderPipelineDesc.multisample.mask = 0xFFFFFFFF;
+    renderPipelineDesc.multisample.alphaToCoverageEnabled = false;
+    // renderPipelineDesc.fragment
+    renderPipelineDesc.fragment = &fragmentState;
+    mRenderPipeline = wgpuDeviceCreateRenderPipeline(device, &renderPipelineDesc);
+    assert(mRenderPipeline);
+}
+
+void WgPipelineRadial::release() {
+    wgpuRenderPipelineRelease(mRenderPipeline);
+    wgpuShaderModuleRelease(mShaderModule);
+    wgpuPipelineLayoutRelease(mPipelineLayout);
+    wgpuBindGroupLayoutRelease(mBindGroupLayout);
+}
diff --git a/src/renderer/wg_engine/tvgWgPipelineRadial.h b/src/renderer/wg_engine/tvgWgPipelineRadial.h
new file mode 100644
index 00000000..f04e12c3
--- /dev/null
+++ b/src/renderer/wg_engine/tvgWgPipelineRadial.h
@@ -0,0 +1,61 @@
+/*
+ * Copyright (c) 2023 the ThorVG project. All rights reserved.
+
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef _TVG_WG_PIPELINE_RADIAL_H_
+#define _TVG_WG_PIPELINE_RADIAL_H_
+
+#include "tvgWgPipelineBase.h"
+
+class WgPipelineRadial;
+
+#define MAX_RADIAL_GRADIENT_STOPS 4
+struct WgPipelineRadialGradientInfo {
+    alignas(16) float nStops[4]{};
+    alignas(16) float centerPos[2]{};
+    alignas(8)  float radius[2]{};
+    alignas(8)  float stopPoints[MAX_RADIAL_GRADIENT_STOPS]{};
+    alignas(16) float stopColors[4 * MAX_RADIAL_GRADIENT_STOPS]{};
+};
+
+struct WgPipelineDataRadial: WgPipelineData {
+    WgPipelineRadialGradientInfo uGradientInfo{}; // @binding(1)
+
+    void updateGradient(RadialGradient* radialGradient);
+};
+
+class WgPipelineBindGroupRadial: public WgPipelineBindGroup {
+private:
+    WGPUBuffer uBufferGradientInfo{}; // @binding(1)
+public:
+    void initialize(WGPUDevice device, WgPipelineRadial& pipelinePipelineRadial);
+    void release();
+
+    void update(WGPUQueue mQueue, WgPipelineDataRadial& pipelineDataRadial);
+};
+
+class WgPipelineRadial: public WgPipelineBase {
+public:
+    void initialize(WGPUDevice device) override;
+    void release() override;
+};
+
+#endif //_TVG_WG_PIPELINE_RADIAL_H_
diff --git a/src/renderer/wg_engine/tvgWgRenderData.cpp b/src/renderer/wg_engine/tvgWgRenderData.cpp
index e0ea6e0e..dc248e3b 100644
--- a/src/renderer/wg_engine/tvgWgRenderData.cpp
+++ b/src/renderer/wg_engine/tvgWgRenderData.cpp
@@ -81,6 +81,8 @@ void WgGeometryData::release() {
 void WgRenderDataShape::release() {
     releaseRenderData();
     mPipelineBindGroupSolid.release();
+    mPipelineBindGroupLinear.release();
+    mPipelineBindGroupRadial.release();
 }
 
 void WgRenderDataShape::releaseRenderData() {
diff --git a/src/renderer/wg_engine/tvgWgRenderData.h b/src/renderer/wg_engine/tvgWgRenderData.h
index c5e304f3..80250dcf 100644
--- a/src/renderer/wg_engine/tvgWgRenderData.h
+++ b/src/renderer/wg_engine/tvgWgRenderData.h
@@ -22,6 +22,7 @@
 
 #include "tvgWgPipelineSolid.h"
 #include "tvgWgPipelineLinear.h"
+#include "tvgWgPipelineRadial.h"
 
 #ifndef _TVG_WG_RENDER_DATA_H_
 #define _TVG_WG_RENDER_DATA_H_
@@ -55,6 +56,7 @@ public:
 
     WgPipelineBindGroupSolid mPipelineBindGroupSolid{};
     WgPipelineBindGroupLinear mPipelineBindGroupLinear{};
+    WgPipelineBindGroupRadial mPipelineBindGroupRadial{};
 
     WgPipelineBase* mPipelineBase{}; // external
     WgPipelineBindGroup* mPipelineBindGroup{}; // external
diff --git a/src/renderer/wg_engine/tvgWgRenderer.cpp b/src/renderer/wg_engine/tvgWgRenderer.cpp
index eb3543c8..49fe8c32 100644
--- a/src/renderer/wg_engine/tvgWgRenderer.cpp
+++ b/src/renderer/wg_engine/tvgWgRenderer.cpp
@@ -104,6 +104,7 @@ void WgRenderer::initialize() {
     mPipelineEmpty.initialize(mDevice);
     mPipelineSolid.initialize(mDevice);
     mPipelineLinear.initialize(mDevice);
+    mPipelineRadial.initialize(mDevice);
     mPipelineBindGroupEmpty.initialize(mDevice, mPipelineEmpty);
     mGeometryDataPipeline.initialize(mDevice);
 }
@@ -118,6 +119,7 @@ void WgRenderer::release() {
     if (mSurface) wgpuSurfaceRelease(mSurface);
     mGeometryDataPipeline.release();
     mPipelineBindGroupEmpty.release();
+    mPipelineRadial.release();
     mPipelineLinear.release();
     mPipelineSolid.release();
     mPipelineEmpty.release();
@@ -137,6 +139,7 @@ RenderData WgRenderer::prepare(const RenderShape& rshape, RenderData data, const
         renderDataShape->initialize(mDevice);
         renderDataShape->mPipelineBindGroupSolid.initialize(mDevice, mPipelineSolid);
         renderDataShape->mPipelineBindGroupLinear.initialize(mDevice, mPipelineLinear);
+        renderDataShape->mPipelineBindGroupRadial.initialize(mDevice, mPipelineRadial);
     }
     
     if (flags & RenderUpdateFlag::Path)
@@ -152,6 +155,14 @@ RenderData WgRenderer::prepare(const RenderShape& rshape, RenderData data, const
             renderDataShape->mPipelineBindGroupLinear.update(mQueue, brushDataLinear);
             renderDataShape->mPipelineBindGroup = &renderDataShape->mPipelineBindGroupLinear;
             renderDataShape->mPipelineBase = &mPipelineLinear;
+        } // setup radial fill properties
+        else if (rshape.fill->identifier() == TVG_CLASS_ID_RADIAL) {
+            WgPipelineDataRadial brushDataRadial{};
+            brushDataRadial.updateMatrix(mViewMatrix, transform);
+            brushDataRadial.updateGradient((RadialGradient*)rshape.fill);
+            renderDataShape->mPipelineBindGroupRadial.update(mQueue, brushDataRadial);
+            renderDataShape->mPipelineBindGroup = &renderDataShape->mPipelineBindGroupRadial;
+            renderDataShape->mPipelineBase = &mPipelineRadial;
         }
     }
 
diff --git a/src/renderer/wg_engine/tvgWgRenderer.h b/src/renderer/wg_engine/tvgWgRenderer.h
index 9c9a6acb..1461591c 100644
--- a/src/renderer/wg_engine/tvgWgRenderer.h
+++ b/src/renderer/wg_engine/tvgWgRenderer.h
@@ -80,6 +80,7 @@ private:
     WgPipelineEmpty mPipelineEmpty;
     WgPipelineSolid mPipelineSolid;
     WgPipelineLinear mPipelineLinear;
+    WgPipelineRadial mPipelineRadial;
     WgGeometryData mGeometryDataPipeline;
     WgPipelineBindGroupEmpty mPipelineBindGroupEmpty;
 };
diff --git a/src/renderer/wg_engine/tvgWgShaderSrc.cpp b/src/renderer/wg_engine/tvgWgShaderSrc.cpp
index 6a2b5e3d..f26d383f 100644
--- a/src/renderer/wg_engine/tvgWgShaderSrc.cpp
+++ b/src/renderer/wg_engine/tvgWgShaderSrc.cpp
@@ -26,6 +26,7 @@
 //************************************************************************
 // cShaderSource_PipelineEmpty
 //************************************************************************
+
 const char* cShaderSource_PipelineEmpty = R"(
 // vertex input
 struct VertexInput {
@@ -61,6 +62,7 @@ fn fs_main(in: VertexOutput) -> void {
 //************************************************************************
 // cShaderSource_PipelineSolid
 //************************************************************************
+
 const char* cShaderSource_PipelineSolid = R"(
 // vertex input
 struct VertexInput {
@@ -102,8 +104,9 @@ fn fs_main(in: VertexOutput) -> @location(0) vec4f {
 })";
 
 //************************************************************************
-// cShaderSource_PipelineRadial
+// cShaderSource_PipelineLinear
 //************************************************************************
+
 const char* cShaderSource_PipelineLinear = R"(
 // vertex input
 struct VertexInput {
@@ -184,3 +187,82 @@ fn fs_main(in: VertexOutput) -> @location(0) vec4f {
 
     return color;
 })";
+
+//************************************************************************
+// cShaderSource_PipelineRadial
+//************************************************************************
+
+const char* cShaderSource_PipelineRadial = R"(
+// vertex input
+struct VertexInput {
+    @location(0) position: vec3f
+};
+
+// Matrix
+struct Matrix {
+    transform: mat4x4f
+};
+
+// GradientInfo
+const MAX_STOP_COUNT = 4;
+struct GradientInfo {
+    nStops     : vec4f,
+    centerPos  : vec2f,
+    radius     : vec2f,
+    stopPoints : vec4f,
+    stopColors : array<vec4f, MAX_STOP_COUNT>
+};
+
+// vertex output
+struct VertexOutput {
+    @builtin(position) position: vec4f,
+    @location(0) vScreenCoord: vec2f
+};
+
+// uMatrix
+@group(0) @binding(0) var<uniform> uMatrix: Matrix;
+// uGradientInfo
+@group(0) @binding(1) var<uniform> uGradientInfo: GradientInfo;
+
+@vertex
+fn vs_main(in: VertexInput) -> VertexOutput {
+    // fill output
+    var out: VertexOutput;
+    out.position = uMatrix.transform * vec4f(in.position.xy, 0.0, 1.0);
+    out.vScreenCoord = in.position.xy;
+    return out;
+}
+
+@fragment
+fn fs_main(in: VertexOutput) -> @location(0) vec4f {
+    // get interpolation factor
+    let t: f32 = clamp(distance(uGradientInfo.centerPos, in.vScreenCoord) / uGradientInfo.radius.x, 0.0, 1.0);
+
+    // get stops count
+    let last: i32 = i32(uGradientInfo.nStops[0]) - 1;
+
+    // resulting color
+    var color = vec4(1.0);
+
+    // closer than first stop
+    if (t <= uGradientInfo.stopPoints[0]) {
+        color = uGradientInfo.stopColors[0];
+    }
+    
+    // further than last stop
+    if (t >= uGradientInfo.stopPoints[last]) {
+        color = uGradientInfo.stopColors[last];
+    }
+
+    // look in the middle
+    for (var i = 0i; i < last; i++) {
+        let strt = uGradientInfo.stopPoints[i];
+        let stop = uGradientInfo.stopPoints[i+1];
+        if ((t > strt) && (t < stop)) {
+            let step: f32 = (t - strt) / (stop - strt);
+            color = mix(uGradientInfo.stopColors[i], uGradientInfo.stopColors[i+1], step);
+        }
+    }
+
+    return color;
+})";
diff --git a/src/renderer/wg_engine/tvgWgShaderSrc.h b/src/renderer/wg_engine/tvgWgShaderSrc.h
index e8db9d92..2707a4cd 100644
--- a/src/renderer/wg_engine/tvgWgShaderSrc.h
+++ b/src/renderer/wg_engine/tvgWgShaderSrc.h
@@ -34,4 +34,7 @@ extern const char* cShaderSource_PipelineSolid;
 // pipeline shader module linear
 extern const char* cShaderSource_PipelineLinear;
 
-#endif
+// pipeline shader module radial
+extern const char* cShaderSource_PipelineRadial;
+
+#endif // _TVG_WG_SHADER_SRC_H_