Hayamimi

Facts

Announced: real-time multilingual speech recognition system that runs on CPU without GPUs or cloud APIs · 2026-10-11
Noted: starts showing subtitles while speech is in progress and finalizes Japanese output about 100ms after a speaker stops · 2026-10-11
Noted: works under 2GB of memory, supports live subtitles, speaker labeling, and translated subtitles · 2026-10-11
Noted: achieves 3.8% character error rate on Japanese broadcast audio · 2026-10-11

Structured graph also available as JSON at /public/entities/hayamimi. CC BY 4.0.

All coverage

2h ago

Hayamimi Runs Real-Time Multilingual Speech Recognition on CPU Alone

Hayamimi is a real-time multilingual speech recognition system that runs on CPU without GPUs or cloud APIs. It starts showing subtitles while speech is in progress and finalizes Japanese output about 100ms after a speaker stops. It works under 2GB of memory, supports live subtitles, speaker labeling, and translated subtitles, and uses INT8-quantized ONNX models on sherpa-onnx.