SenseVoice is a speech foundation model from Alibaba's FunAudioLLM team that delivers multiple speech understanding capabilities including automatic speech recognition (ASR), spoken language identification (LID), speech emotion recognition (SER), and audio event detection (AED). Trained with over 400,000 hours of data and supporting more than 50 languages, it surpasses Whisper in recognition performance while running 7x faster than Whisper-small and 17x faster than Whisper-large. The model family includes SenseVoice-Small for low-latency 5-language ASR and SenseVoice-Large for high-precision 50+ language support.