中
◀ Back to feed
TechCrunch

Underdog shows off a 32MB speech model in new demo

Published 2026-11-06 Collected 2026-11-07 00:00 AI news #ai#tech
Full text

AI companies across the world are racing to make bigger and better models, but the problem is all of them need massive amounts of processing power. On the other end, some researchers and startups are creating smaller models that can run on edge devices, like your phone or computer.

One such startup, Underdog, has released a game built with its new voice model called Woof that's only 32MB large.

The game features a room with a retro tech theme, a typewriter, and a CRT monitor. There is a desk lamp, which makes a sound when you click its switch. The intro warns users about "intrusive" cloud providers in the town — ostensibly a jab at cloud-based models.

You can type anything on the typewriter using your keyboard, and the game will speak it back. You can also rotate the tuner to generate different kinds of music.

All of this works within the browser through Underdog's speech generation model, which, the company says, can run on most devices without any internet connection. The model has 31 million parameters and is open for everyone to download.

While such a small model likely won't replace bigger models, it may still be enough to drive local assistant experiences or create personalized podcasts.

This demo comes weeks after Underdog launched Bark, a 167 MB speech-to-text model also designed to run on edge devices for transcription and dictation.

Sigil Wen, the founder of Underdog, told TechCrunch that he started building the app because, instead of using notetakers like Granola, he wanted to build his own system with AI models that can run on users' devices and let them own their data.

"I think we will be able to run most capable models of today on the devices you already own within months. We just want to train the best open models that are American-made, that are open, that are capable and allow people to have private AI," he said, referring to his manifesto.

In addition to the two voice models mentioned above, Underdog has also released a 27-billion-parameter model that can run on laptops, along with a waitlisted app that the company is positioning as a locally-hosted alternative to Muse and Instinct.

Source
Title
Underdog shows off a 32MB speech model in new demo
Author
Ivan Mehta
Site
TechCrunch
Published
2026-11-06
Collected
2026-11-07 00:00 · 2026-11-07-00-ai.json
Link
https://techcrunch.com/2026/11/06/underdog-shows-off-a-32mb-speech-model-in-new-demo/
Open original ↗

Reposted. Will be removed on request. All rights belong to the original author.

More from this batch

00:00batch · 2026-11-0712 items
>

Ramp data shows AI spending stalling just as Anthropic nears its IPO

Ramp reports token spend grew only 1.5% across August–September while usage rose, as price cuts and cost controls cap vendor revenue ahead o…

TechCrunch
>

Anthropic's $1.5 trillion IPO hits resistance as investors balk at valuation

Investors push back on Anthropic's $1.5T IPO valuation amid tech selloff and missing October revenue in the S-1; a delay could shake the bro…

The Decoder
>

Frontier Radar #5:若 AGI 已近,为何大佬还在出走办「新实验室」?梳理 45 家 neolab 的六类技术赌注

专题把独立研究型 AI 公司称为 neolab,按「更好的学习方式 / 不只靠语言 / 独占真实数据 / 算力效率 / 自动化科研 / 不拼第一名也赚钱」六类赌注整理约 27 家详档加近 20 家简述,并点出英伟达几乎处处下注。适合做战略扫描:看清大厂产品化压力之外,资本在押哪条…

The Decoder
>

Stratechery Sharp Tech 上视频:美国选民如何投票「数据中心」、对大科技与 Google AI 回应意味着什么

Ben Thompson 把 Sharp Tech 首次做成视频,讨论中期选举里数据中心议题如何影响大科技叙事,以及 Google 对 AI 竞争的回应。偏宏观产业与资本开支政治风险,适合对照本周公投与抛售新闻。

Stratechery
>

After blocking Muse, Amazon now lets users text Alexa to help them shop

Amazon is rolling out texting with Alexa for Shopping so U.S. customers can search, compare and buy over SMS, after previously blocking Meta…

Amazon / TechCrunch
>

Visa Agentic Commerce Benchmark: almost half of AI shopping test runs fail

Visa's first agentic shopping benchmark scores 22 models at a 57.2% pass rate; 54% of failed runs still ended in a purchase.

Visa
>

Google Antigravity adds remote agents and voice mode

Antigravity v2.29 starts rolling out cloud remote agents for AI Pro/Ultra and Gemini Live voice mode, plus redesigned Automations and a nati…

Google Antigravity
>

Xiaohongshu's Mishu Personal Agent launches via Rokid AI glasses

Xiaohongshu frames Mishu as a Personal Agent and pairs it with Rokid glasses for progressive on-site sensing grounded in community notes.

界面新闻
>

China Mobile open-sources JoinAI-Agent engine

China Mobile released Jiutian Juzhi JoinAI-Agent, a multi-agent engine that extends MCP tools and A2A agents with zero-code workflows.

中国移动
>

Physicl raises $10M to make the digital world real enough for robots

Physicl's Studio turns high-fidelity 3D pipelines into synthetic data tooling for robot and world-model post-training.

TechCrunch
>

ToyBlocks: A controlled world for robot learning and evaluation

Ai2's first FoundationWorld is an open tabletop block-manipulation sim with automatic demos, metrics, and early sim-to-real transfer results…

Hugging Face / Ai2
>

Nace.AI 开源 Drex 决策模型族:输入状态 + 题型,一次前向给出各选项概率,不对文本「生成」

改进什么:很多路由/风控/分类只要校准概率,不必生成长文。怎么做:Drex Max(27B 多模态)、Drex v2.0(35B MoE/约 3B 激活)等,兼容 TypeSafe 风格 `/v1/systemone`,可本地 Python/llama.cpp/Ollama。许可…

Nace.AI