Research

Open-source cross-platform mobile app for fully offline speech translation — combining on-device ASR (SenseVoice), neural machine translation, and TTS on iOS and Android with system audio capture

Benchmarking bidirectional English-Japanese speech translation models — Qwen3-ASR (1.7B, highest quality) vs distilled Whisper (756M, 4x faster) — against OpenAI Whisper large-v3 and Meta SeamlessM4T v2

Comprehensive benchmark of 16 on-device speech-to-text models across 9 inference engines on Android, iOS, macOS, and Windows

Comprehensive benchmark of 18 on-device text-to-speech models including Kokoro, Piper, Matcha, Kitten, and VITS on Android and iOS

How VoicePing engineered Bilingual Mode for automatic, low-latency language switching inside a single WebSocket stream powered by customized Whisper V2 models.

Technical comparison of NeMo MSDD and Pyannote 3.1 across 6 real-world test scenarios