NetEase Youdao open-sources Confucius4-R2T2, a stable-prefix streaming ASR with 80ms chunks
pmttyji · reddit · 2026-11-09
NetEase Youdao released Confucius4-R2T2, a low-latency stable-prefix streaming ASR model for "real real-time" transcription.
- Configurable decoding chunks from 80ms to 2s; monotonic output commits text permanently with no revisions, reducing flicker in live captioning, simultaneous translation, and LLM agent pipelines
- Adapts Qwen3-ASR-1.7B with Longest Stable Prefix (LSP) learning on 1,500 hours of Chinese/English audio; two-stage SFT + GRPO training updates only the language decoder, with an explicit Wait action deciding when to commit
- Code under Apache 2.0, weights under NetEase's own license; GGUF quantization, demo Space, GitHub repo, and arXiv paper all available
Related event: NetEase Youdao Open-Sources Confucius4-R2T2 for Real-Time Streaming ASR(2 posts)→
More from Models
- Claude Fable 5.5 needs a 70 on AA index to justify replacing Opus 5.5 — haider1 · 2026-11-09
- Leaked checkpoints suggest Google now ships a major Gemini release every month — lyraxana · 2026-11-09
- HF's Niels Rogge releases 'Getting Started with Local AI' slides: quantization, MoE, local agents — NielsRogge · 2026-11-09
- InstaDeep CEO: Anthropic's unreleased Models 3 and 4 sit halfway to AI2027's Agent-2 — connoraxiotes · 2026-11-09
- 27B open-weight decision model Drex Max takes #1 on Artificial Analysis Decision Index — dr_cintas · 2026-11-09
- JEV models get full support in Hugging Face transformers, now standardized — art_zucker · 2026-11-09