ZERO SERVER LOAD • WEBGPU & WASM

Edge Benchmarking Lab

Stress-test .ONNX vision models directly on your client hardware (WebGPU acceleration or multi-threaded CPU SIMD).

DETECTED CLIENT HARDWARE

Detecting Hardware... (8 Cores)

1. Configure Edge Execution

Model Source (.onnx or instant synthetic test)⚡ Built-in YOLO11n Tensor Benchmark (Or click to upload .onnx)Browse ONNX

Execution Engine Provider

Input Resolution

Timed Iterations

Ready for Client-Side Telemetry

Click Run Local Hardware Benchmark to measure your device's exact sustained FPS, warm-up overhead, and P95 tail latency against Vyrion Cloud GPUs.