diff --git a/gpu/src/Limits.zig b/gpu/src/Limits.zig
index 055b605a..a565d875 100644
--- a/gpu/src/Limits.zig
+++ b/gpu/src/Limits.zig
@@ -1,37 +1,27 @@
 // TODO: docs
-
-// TODO:
-// typedef struct WGPUSupportedLimits {
-//     WGPUChainedStructOut * nextInChain;
-//     WGPULimits limits;
-// } WGPUSupportedLimits;
-
-// TODO:
-// typedef struct WGPULimits {
-//     uint32_t maxTextureDimension1D;
-//     uint32_t maxTextureDimension2D;
-//     uint32_t maxTextureDimension3D;
-//     uint32_t maxTextureArrayLayers;
-//     uint32_t maxBindGroups;
-//     uint32_t maxDynamicUniformBuffersPerPipelineLayout;
-//     uint32_t maxDynamicStorageBuffersPerPipelineLayout;
-//     uint32_t maxSampledTexturesPerShaderStage;
-//     uint32_t maxSamplersPerShaderStage;
-//     uint32_t maxStorageBuffersPerShaderStage;
-//     uint32_t maxStorageTexturesPerShaderStage;
-//     uint32_t maxUniformBuffersPerShaderStage;
-//     uint64_t maxUniformBufferBindingSize;
-//     uint64_t maxStorageBufferBindingSize;
-//     uint32_t minUniformBufferOffsetAlignment;
-//     uint32_t minStorageBufferOffsetAlignment;
-//     uint32_t maxVertexBuffers;
-//     uint32_t maxVertexAttributes;
-//     uint32_t maxVertexBufferArrayStride;
-//     uint32_t maxInterStageShaderComponents;
-//     uint32_t maxComputeWorkgroupStorageSize;
-//     uint32_t maxComputeInvocationsPerWorkgroup;
-//     uint32_t maxComputeWorkgroupSizeX;
-//     uint32_t maxComputeWorkgroupSizeY;
-//     uint32_t maxComputeWorkgroupSizeZ;
-//     uint32_t maxComputeWorkgroupsPerDimension;
-// } WGPULimits;
+max_texture_dimension_1d: u32,
+max_texture_dimension_2d: u32,
+max_texture_dimension_3d: u32,
+max_texture_array_layers: u32,
+max_bind_groups: u32,
+max_dynamic_uniform_buffers_per_pipeline_layout: u32,
+max_dynamic_storage_buffers_per_pipeline_layout: u32,
+max_sampled_textures_per_shader_stage: u32,
+max_samplers_per_shader_stage: u32,
+max_storage_buffers_per_shader_stage: u32,
+max_storage_textures_per_shader_stage: u32,
+max_uniform_buffers_per_shader_stage: u32,
+max_uniform_buffer_binding_size: u64,
+max_storage_buffer_binding_size: u64,
+min_uniform_buffer_offset_alignment: u32,
+min_storage_buffer_offset_alignment: u32,
+max_vertex_buffers: u32,
+max_vertex_attributes: u32,
+max_vertex_buffer_array_stride: u32,
+max_inter_stage_shader_components: u32,
+max_compute_workgroup_storage_size: u32,
+max_compute_invocations_per_workgroup: u32,
+max_compute_workgroup_size_x: u32,
+max_compute_workgroup_size_y: u32,
+max_compute_workgroup_size_z: u32,
+max_compute_workgroups_per_dimension: u32,
diff --git a/gpu/src/NativeInstance.zig b/gpu/src/NativeInstance.zig
index 87bec0cf..079053e3 100644
--- a/gpu/src/NativeInstance.zig
+++ b/gpu/src/NativeInstance.zig
@@ -300,10 +300,36 @@ const device_vtable = Device.VTable{
     }).release,
 };
 
+// TODO: maybe make Limits an extern struct that can be cast?
 fn convertLimits(l: Limits) c.WGPULimits {
-    // TODO: implement!
-    _ = l;
-    return std.mem.zeroes(c.WGPULimits);
+    return .{
+        .maxTextureDimension1D = l.max_texture_dimension_1d,
+        .maxTextureDimension2D = l.max_texture_dimension_2d,
+        .maxTextureDimension3D = l.max_texture_dimension_3d,
+        .maxTextureArrayLayers = l.max_texture_array_layers,
+        .maxBindGroups = l.max_bind_groups,
+        .maxDynamicUniformBuffersPerPipelineLayout = l.max_dynamic_uniform_buffers_per_pipeline_layout,
+        .maxDynamicStorageBuffersPerPipelineLayout = l.max_dynamic_storage_buffers_per_pipeline_layout,
+        .maxSampledTexturesPerShaderStage = l.max_sampled_textures_per_shader_stage,
+        .maxSamplersPerShaderStage = l.max_samplers_per_shader_stage,
+        .maxStorageBuffersPerShaderStage = l.max_storage_buffers_per_shader_stage,
+        .maxStorageTexturesPerShaderStage = l.max_storage_textures_per_shader_stage,
+        .maxUniformBuffersPerShaderStage = l.max_uniform_buffers_per_shader_stage,
+        .maxUniformBufferBindingSize = l.max_uniform_buffer_binding_size,
+        .maxStorageBufferBindingSize = l.max_storage_buffer_binding_size,
+        .minUniformBufferOffsetAlignment = l.min_uniform_buffer_offset_alignment,
+        .minStorageBufferOffsetAlignment = l.min_storage_buffer_offset_alignment,
+        .maxVertexBuffers = l.max_vertex_buffers,
+        .maxVertexAttributes = l.max_vertex_attributes,
+        .maxVertexBufferArrayStride = l.max_vertex_buffer_array_stride,
+        .maxInterStageShaderComponents = l.max_inter_stage_shader_components,
+        .maxComputeWorkgroupStorageSize = l.max_compute_workgroup_storage_size,
+        .maxComputeInvocationsPerWorkgroup = l.max_compute_invocations_per_workgroup,
+        .maxComputeWorkgroupSizeX = l.max_compute_workgroup_size_x,
+        .maxComputeWorkgroupSizeY = l.max_compute_workgroup_size_y,
+        .maxComputeWorkgroupSizeZ = l.max_compute_workgroup_size_z,
+        .maxComputeWorkgroupsPerDimension = l.max_compute_workgroups_per_dimension,
+    };
 }
 
 test "syntax" {
diff --git a/gpu/src/TODO b/gpu/src/TODO
index 9eeeef96..1bc8f1ba 100644
--- a/gpu/src/TODO
+++ b/gpu/src/TODO
@@ -994,6 +994,11 @@ typedef struct WGPUComputePipelineDescriptor {
     WGPUProgrammableStageDescriptor compute;
 } WGPUComputePipelineDescriptor;
 
+typedef struct WGPUSupportedLimits {
+    WGPUChainedStructOut * nextInChain;
+    WGPULimits limits;
+} WGPUSupportedLimits;
+
 typedef struct WGPUDeviceProperties {
     uint32_t deviceID;
     uint32_t vendorID;