GoKawiilUnderdog releases 32MB text-to-speech model Woof via browser game demo
Startup Underdog launched Woof, a 31.7MB text-to-speech model with 31 million parameters that runs entirely in a browser without internet access. The company showcased it through a retro-themed browser game where users can type text for the model to speak and manipulate a music tuner, all voiced by Woof. Underdog made the model open for download, and the demo follows the company's earlier release of Bark, a 167MB speech-to-text model for edge devices.
GoKawiil's interpretation of the reporting above, not reported fact.
The release reflects a broader industry split between companies building ever-larger AI models and those optimizing for small, on-device models that need no cloud connection. Underdog founder Sigil Wen has said he wants to build AI tools that run locally so users retain control of their data, suggesting a philosophy centered on privacy and offline capability rather than raw model power. Such compact models are unlikely to rival large cloud-based systems in capability, but could enable local voice assistants or personalized audio tools without server dependency.
- Underdog's Woof is a 31.7MB, 31-million-parameter text-to-speech model that runs in-browser offline.
- The company demonstrated it via an interactive retro-themed game rather than a standard product launch.
- It follows Underdog's earlier 167MB speech-to-text model Bark, continuing a focus on edge-device AI.
Source: techcrunch.com — Ivan Mehta, 2026-11-06
Published there as: “Underdog shows off a 32MB speech model in new demo”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.