Two of the most-watched home robots of 2026 tied their credibility to the same season. Weave Robotics launched Isaac 1 on July 1 and says California deliveries begin this fall. Sunday, which raised $165 million in March at a $1.15 billion valuation under the banner "No More Demos," says its Memo robot will enter a beta program placing units in real homes this fall too. Look closely, though, and the two promises are not equivalent.
Isaac 1 — Weave Robotics
Memo — Sunday Robotics (ACT-2 preview)
Isaac 1 is a wheeled robot that collapses to about 90 cm and extends to roughly 1.75 m, with two arms and swappable fabric shells in five colourways. Weave, founded by Kaan Dogrusoz and Evan Wineland, says it handles laundry and daily tidying autonomously by default, with a remote human teleoperator stepping in — able to see into the home through the robot's cameras — when the robot hits a task it can't complete alone. Weave says the cameras physically power off when not in use. It's already taking orders: a refundable $250 deposit, then $7,999 upfront or $449/month at delivery. The order terms are worth reading past the marketing page, though — Weave's own agreement states a deposit reserves a place in line but does not obligate the company to deliver a device, and that specifications and timing may still change.
Memo works differently. Instead of live human teleoperation during use, Sunday trained its ACT-1 model beforehand on roughly 10 million household task recordings, captured by volunteer "Memory Developers" wearing a sensor-equipped glove in more than 500 homes. Reported specs put Memo at about 170 cm extended, 77 kg, with roughly 4 hours of runtime per charge. Sunday's Founding Family Beta plans to place 50 units in homes later this year. Memo is not yet open for pre-order the way Isaac 1 is, and industry analysts have flagged a real gap between Sunday's target retail price (under $10,000) and what a unit reportedly costs to build today (around $20,000) — a gap that would need to close before broad availability means anything.
That difference matters for how these two promises should be read. Isaac 1's open question is about the boundaries of live human oversight inside your home. Memo's open question is whether a model trained on recorded demonstrations generalises reliably once it's making decisions on its own, and whether the unit economics hold up outside a subsidised beta. Treating them as the same kind of "fall 2026" promise would blur two genuinely different bets.
What would count as real evidence, rather than a company-controlled demo, for either one: independent accounts from an actual test household, precise disclosure of how often (not just whether) a human intervenes, and the details announcements tend to skip — reliability across repeated use, and what happens when a task fails.
DOMULU cannot run its own in-home test yet. But this is exactly the kind of promise-versus-delivery moment worth tracking, and we'll follow up as verifiable details actually surface — not when a press release lands.
This article is a research preview. Unless explicitly stated, it does not report a completed independent product test by DOMULU.