NetEase Youdao Open-Sources Confucius R2T2: Streaming ASR with 80ms Decode Chunks
VraserX · x · 2026-09-17
NetEase Youdao open-sourced Confucius4-R2T2, a low-latency streaming ASR model on GitHub.
- The poster cautions that the headline 160ms figure is the minimum streaming audio step, not end-to-end response latency — an important distinction
- Decoding chunks are fine-grained and configurable from 80ms to 2s
- Append-only output mode: committed transcript text is never revised, avoiding text flickering in real-time applications
- Target uses: live captioning, downstream NLP pipelines and LLM agents, simultaneous translation
- Repo ships example scripts plus WebSocket server/client code
Related event: NetEase Youdao Open-Sources Streaming ASR Model Confucius4-R2T2(21 posts)→
More from Models
- Chinese AI labs fear US winning recursive self-improve as distillation gets harder — McDonaghMatthew · 2026-11-03
- Editing 0.016% of weights fixes Qwen reasoning loops, MATH-500 jumps from 7.9% to 48.1% — DeliciousGorilla · 2026-11-03
- Did OpenAI rush its math-breakthrough model past safety testing? — haider1 · 2026-11-03
- Don't trust viral GPT-6.2 Astra demos: PS5 controller SVG exposes Python files behind the image — Angaisb_ · 2026-11-03
- Llama.app adds Liquid AI LFM3, Mistral Large 4 and coding-agent model Flint — mervenoyann · 2026-11-03
- ChatGPT/Codex 5-hour usage limits appear to have kicked in earlier than December — azed_ai · 2026-11-03