Investors

The short pitch.

MingLLM is a personal-intelligence runtime — voice, browser, research, and receipts — that runs entirely on the user's device. We're raising a seed round to ship v1 and grow the team from one to eight. If this fits your thesis, we'd like to talk.

4
Product surfaces · voice, browser, coding, answers
1
Peer-style paper published · Mneme memory
1→8
Team size · current → post-round
0
Cloud calls by default
01

Local is the moat

Frontier compute is a liability when your product lives on the user's device. We chose that constraint — our training loop, our inference target, and our product shape all compound on that decision.

02

Bounded agency wins first

Open-ended general intelligence is a decade problem. Bounded agency for a single user is a this-year problem. We ship the second and fund the first with it.

03

Receipts are a feature, not overhead

Every action MingLLM takes is logged and reviewable. This is not compliance theatre — it's what makes users trust the agent enough to give it real authority.

Get in touch

Deck + financial model on request.

Email the founder directly. Please include your fund's thesis and typical check size so we can respond with the right materials.

yimingbeckmann@mingllm.com