We’re the team behind Own 1, and we’d like to hear from people who run local models today. Own 1 runs two local LLMs and eight agents on a fanless machine, with no subscriptions and no per-token cost.

A few questions we’d love your take on:

1. What workload would justify buying one? Private document search, coding agents, a shared household or studio assistant, or something else?

2. Where do you draw the line on cloud use? Own 1 stays on-device by default, bursts to OwnXCloud only under your rules, and never calls a hosted API without your approval. Does that match how you’d want it to work?

3. Would you let an agent spend money? Agents can pay for approved models and compute with USDC through x402, under per-agent budgets and approval thresholds. What limits would you set?

4. What’s missing? Models, ports, apps, or anything else that would stop you from switching from a Mac Studio or a Strix Halo mini PC?

Specs are preliminary until launch, so your feedback can still change what we ship. Ask us anything, including the hard questions about benchmarks and limits.