NetEase Youdao open-sources Confucius4-R2T2 with tunable 80ms-2s decoding chunks
aigclink · x · 2026-09-20
NetEase Youdao has open-sourced Confucius4-R2T2, a low-latency, high-accuracy true streaming ASR model.
- Real-time, append-only: fine-grained decoding chunks configurable from 80ms to 2s; committed transcript text is never revised, avoiding flickering in live captioning.
- Use cases: live subtitling, simultaneous speech translation, voice customer service, and downstream NLP pipelines or LLM agents that must act on text instantly.
- Repo: ships with example.py, WebSocket server/client scripts, and bilingual docs for quick deployment.
Related event: NetEase Youdao Open-Sources Confucius4-R2T2 Real-Time ASR(2 posts)→
More from Models
- Sentence Transformers models quietly dominate Hugging Face's most-downloaded list — tomaarsen · 2026-09-20
- Astra aces the pelican test in Nautilo: co-creative design without prompt-and-pray — Dan_Jeffries1 · 2026-09-20
- Model negativity trends upward since Opus 3, which had lowest self-distress — repligate · 2026-09-20
- DeepMind exec 100% certain of frontier return as Gemini 4 slips — nathanbenaich · 2026-09-20
- Why strong open models matter: fine-tuning goes a long way, zero-shot buys flexibility — antoine_chaffin · 2026-09-20
- DiffusionGemma as Jev: single-pass parallel denoising decisions in ~0.2s on DGX Spark — bodonoghue85 · 2026-09-20