Underdog Woof: 31.7MB open-weight TTS model claims 1.42% WER, runs fully in-browser
_sathvikr · x · 2026-11-06
Underdog released Woof, an open-weight (CC BY-NC 4.0) text-to-speech model at just 30.7M params / 31.7MB that runs entirely in-browser via WebGPU. The team claims a 1.42% word error rate, 1,022x smaller than Fish Audio Drama 3 with 25% better WER, and up to 18x real-time speed on a single MacBook CPU thread. They launched a browser game, Underdog Woof Escape, voiced entirely by Woof, with all audio and music generated offline by local models. The decoder follows Kyutai's Mimi design; training data includes LibriTTS-R.
Related event: Underdog's 31.7MB Open-Source TTS Model Woof Runs Fully in the Browser(2 posts)→
More from Infra
- Kimi K3.1 Pro gets Day-0 TPU support: serving a 2.8T-param MoE with llm-d on TPUs and GPUs — vllm_project · 2026-11-06
- Satirical pitch: cool data centers with fluorinert and market them as 'water-free' — typedfemale · 2026-11-06
- US stocks lost $3.8T in a day; chipmakers took $1.15T of the hit — CharlieMuirhead · 2026-11-06
- 'This is why we need at least a datacenter/home': the case for local compute — tech__unicorn · 2026-11-06
- Reddit hopes for cheap RAM by Christmas as memory prices stay in focus — keepthepace · 2026-11-06
- J.B. Huang's new video explains how LLM serving actually works, from fast to cheap — jbhuang0604 · 2026-11-06