diff --git a/gpu/src/Device.zig b/gpu/src/Device.zig
index 013ae90c..f59ad84d 100644
--- a/gpu/src/Device.zig
+++ b/gpu/src/Device.zig
@@ -12,6 +12,7 @@ const Queue = @import("Queue.zig");
 const ShaderModule = @import("ShaderModule.zig");
 const Surface = @import("Surface.zig");
 const SwapChain = @import("SwapChain.zig");
+const RenderPipeline = @import("RenderPipeline.zig");
 
 const Device = @This();
 
@@ -33,7 +34,7 @@ pub const VTable = struct {
     // WGPU_EXPORT WGPUPipelineLayout wgpuDeviceCreatePipelineLayout(WGPUDevice device, WGPUPipelineLayoutDescriptor const * descriptor);
     // WGPU_EXPORT WGPUQuerySet wgpuDeviceCreateQuerySet(WGPUDevice device, WGPUQuerySetDescriptor const * descriptor);
     // WGPU_EXPORT WGPURenderBundleEncoder wgpuDeviceCreateRenderBundleEncoder(WGPUDevice device, WGPURenderBundleEncoderDescriptor const * descriptor);
-    // WGPU_EXPORT WGPURenderPipeline wgpuDeviceCreateRenderPipeline(WGPUDevice device, WGPURenderPipelineDescriptor const * descriptor);
+    createRenderPipeline: fn (ptr: *anyopaque, descriptor: *const RenderPipeline.Descriptor) RenderPipeline,
     // WGPU_EXPORT void wgpuDeviceCreateRenderPipelineAsync(WGPUDevice device, WGPURenderPipelineDescriptor const * descriptor, WGPUCreateRenderPipelineAsyncCallback callback, void * userdata);
     // WGPU_EXPORT WGPUSampler wgpuDeviceCreateSampler(WGPUDevice device, WGPUSamplerDescriptor const * descriptor);
     createShaderModule: fn (ptr: *anyopaque, descriptor: *const ShaderModule.Descriptor) ShaderModule,
@@ -81,6 +82,10 @@ pub inline fn destroy(device: Device) void {
     device.vtable.destroy(device.ptr);
 }
 
+pub inline fn createRenderPipeline(device: Device, descriptor: *const RenderPipeline.Descriptor) RenderPipeline {
+    return device.vtable.createRenderPipeline(device.ptr, descriptor);
+}
+
 pub const Descriptor = struct {
     label: ?[*:0]const u8 = null,
     required_features: ?[]Feature = null,
@@ -100,6 +105,7 @@ test "syntax" {
     _ = createShaderModule;
     _ = nativeCreateSwapChain;
     _ = destroy;
+    _ = createRenderPipeline;
     _ = Descriptor;
     _ = LostReason;
 }
diff --git a/gpu/src/NativeInstance.zig b/gpu/src/NativeInstance.zig
index b9f61d8f..b9df3782 100644
--- a/gpu/src/NativeInstance.zig
+++ b/gpu/src/NativeInstance.zig
@@ -378,8 +378,68 @@ const device_vtable = Device.VTable{
             c.wgpuDeviceDestroy(@ptrCast(c.WGPUDevice, ptr));
         }
     }).destroy,
+    .createRenderPipeline = (struct {
+        pub fn createRenderPipeline(ptr: *anyopaque, descriptor: *const RenderPipeline.Descriptor) RenderPipeline {
+            const desc = convertRenderPipelineDescriptor(descriptor);
+            return wrapRenderPipeline(c.wgpuDeviceCreateRenderPipeline(@ptrCast(c.WGPUDevice, ptr), &desc));
+        }
+    }).createRenderPipeline,
 };
 
+inline fn convertRenderPipelineDescriptor(d: *const RenderPipeline.Descriptor) c.WGPURenderPipelineDescriptor {
+    return c.WGPURenderPipelineDescriptor{
+        .nextInChain = null,
+        .label = d.label,
+        .layout = @ptrCast(c.WGPUPipelineLayout, d.layout.ptr),
+        .vertex = c.WGPUVertexState{
+            .nextInChain = null,
+            .module = @ptrCast(c.WGPUShaderModule, d.vertex.module.ptr),
+            .entryPoint = d.vertex.entry_point,
+            .constantCount = 0, // d.vertex.constants.len,
+            .constants = null, // TODO: need to convert entire list to WGPUConstantEntry
+            .bufferCount = 0, // d.vertex.buffers.len
+            .buffers = null, // TODO: need to convert entire list to WGPUVertexBufferLayout
+        },
+        .primitive = c.WGPUPrimitiveState{
+            .nextInChain = null,
+            .topology = @enumToInt(d.primitive.topology),
+            .stripIndexFormat = @enumToInt(d.primitive.strip_index_format),
+            .frontFace = @enumToInt(d.primitive.front_face),
+            .cullMode = @enumToInt(d.primitive.cull_mode),
+        },
+        // TODO: don't create ptr on stack or something
+        .depthStencil = &c.WGPUDepthStencilState{
+            .nextInChain = null,
+            .format = @enumToInt(d.depth_stencil.format),
+            .depthWriteEnabled = d.depth_stencil.depth_write_enabled,
+            .depthCompare = @enumToInt(d.depth_stencil.depth_compare),
+            .stencilFront = @bitCast(c.WGPUStencilFaceState, d.depth_stencil.stencil_front),
+            .stencilBack = @bitCast(c.WGPUStencilFaceState, d.depth_stencil.stencil_back),
+            .stencilReadMask = d.depth_stencil.stencil_read_mask,
+            .stencilWriteMask = d.depth_stencil.stencil_write_mask,
+            .depthBias = d.depth_stencil.depth_bias,
+            .depthBiasSlopeScale = d.depth_stencil.depth_bias_slope_scale,
+            .depthBiasClamp = d.depth_stencil.depth_bias_clamp,
+        },
+        .multisample = c.WGPUMultisampleState{
+            .nextInChain = null,
+            .count = d.multisample.count,
+            .mask = d.multisample.mask,
+            .alphaToCoverageEnabled = d.multisample.alpha_to_coverage_enabled,
+        },
+        // TODO: don't create ptr on stack or something
+        .fragment = &c.WGPUFragmentState{
+            .nextInChain = null,
+            .module = @ptrCast(c.WGPUShaderModule, d.fragment.module.ptr),
+            .entryPoint = d.vertex.entry_point,
+            .constantCount = 0, // d.fragment.constants.len,
+            .constants = null, // TODO: need to convert entire list to WGPUConstantEntry
+            .targetCount = 0, // d.fragment.targets.len,
+            .targets = null, // TODO: need to convert entire list to WGPUColorTargetState
+        },
+    };
+}
+
 // TODO: maybe make Limits an extern struct that can be cast?
 fn convertLimits(l: Limits) c.WGPULimits {
     return .{