Benchmarks — Android

RTF (real-time factor) below 1.0 means faster than real-time.

Android (ONNX Runtime)

Measured on Android emulator (arm64-v8a, no NNAPI). Real hardware with NNAPI is significantly faster.

ModelTaskAudioInferenceRTF
Parakeet TDT v3STT (114 languages)1.5s175ms0.12
Kokoro 82MTTS (7 languages)1.9s output1,075ms0.58
Silero VAD v5VAD32ms chunk<1ms<0.01
DeepFilterNet3Noise cancellation32ms chunk~5ms~0.15
PlatformAccelerationChipsets
AndroidNNAPISnapdragon 8 Gen 1+, Exynos 2200+, Google Tensor G2+
Embedded LinuxQNN (Hexagon DSP)SA8295P, SA8255P
AnyCPU (XNNPACK)All arm64-v8a / x86_64
Note

Android benchmarks are from emulator without hardware acceleration. On real Snapdragon hardware with NNAPI delegation, expect 2–3x faster inference. Total model size: ~1.2 GB (INT8 quantized ONNX).