Sigil Wen, a Thiel Fellow, has launched Underdog, an invite-only beta for a new AI assistant. This assistant is designed to operate entirely on the user's device, initially supporting Macs and Windows PCs, with future plans for Linux, iPhone, and Android versions. The core feature of Underdog is its emphasis on user privacy, as all data processing occurs locally.
Underdog is powered by Husky, an inference engine developed by Wen, which is optimized for running AI models quickly on local hardware. Wen states that Husky minimizes data transfer between a computer's main chip and graphics chip compared to other on-device engines. Additionally, Underdog includes security features such as encrypting access keys for authorized email and other accounts.
Underdog utilizes a 27-billion parameter reasoning model, fine-tuned from Qwen3.8-27B. While smaller than current state-of-the-art cloud-hosted models, Wen claims its performance is comparable to models like Claude Opus 4.6 from six months prior in certain benchmarks. This level of capability is intended to handle common AI assistant tasks, such as shopping research or answering homework questions, without requiring users to sacrifice privacy.
The initial business model for Underdog involves offering the app for free, with no plans for ad support. Wen anticipates that the capabilities of small on-device models will continue to improve over time, further enhancing the value proposition of privacy-focused AI assistants like Underdog.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's briefNew every morning, and the back catalogue is archived by date.
Sigil Wen, a Thiel Fellow, launched Underdog, an invite-only beta of an AI assistant that runs entirely on-device for enhanced privacy. Underdog uses Wen's Husky inference engine to run a 27-billion parameter model, aiming to provide capabilities comparable to older state-of-the-art models without compromising user data. This development matters as it offers a privacy-focused alternative in the AI assistant market by keeping user data local.