Available
A buyer can place an order or sign a standard commercial path now.
Methodology
What we will defend.
This MVP uses labeled draft data. Scores are editorial composites from public materials, not lab tests and not vendor audits.
Status tiers
Pilot
Named deployments or partner programs; not a self-serve catalog buy.
Pre-order
Public price or waitlist; delivery is vendor-stated.
Announced
Current-gen spec exists; the unit is not the one operating today.
Limited
Real commercial path, geographically or contractually constrained.
Research / prototype
Cannot be bought or pre-ordered. Not Available, not Pre-order, not Coming soon. Compare for evidence only. Primary CTA is Follow research / keep shortlist — not Request demo.
Evidence score (0–100)
Four buckets. We do not average social-media virality into the score.
Specs published /25
Official, citable numbers vs. secondary blogs.
Independent reporting /25
Press, filings, teardowns not written by the maker.
Deployments /30
Paid hours, named sites, shipment figures — not keynotes.
Safety / support docs /20
Public safety story, service path, manuals.
Digit scores high because of operating hours. U1 scores low because it is launch-week claims. That is a documentation score, not a “better robot” score.
All-in cues
Headline prices lie by omission. We surface commercial path (list, RaaS, quote, rental) and the first hidden lines: integration, safety, freight, hands, people.
Vs pages
A vs page is a use-case verdict strip. It is not a trophy. We will not sell a “winner” module. Sponsored rows, when present, are labeled.
Generations
One URL per model. Current generation is canonical; prior gens are notes. Digit and Figure generations are separate — Figure 03 is not Figure 02 with a badge, and Digit 5 specs are not mixed into Digit 4 hours.
What we will not sell
- Organic catalog order
- Evidence scores
- Use-case verdicts
- Fake scarcity on lead forms
Vendors may buy labeled listings, vs insight rows, and introduced leads. See For vendors.
FAQ
- How do you score humanoid robots?
- Evidence buckets: deployments, third-party records, price transparency, and availability. Marketing claims alone do not raise the score. See this methodology page.
- What does “all-in” mean?
- List price or lease is only the start. Integration, safety validation, ops cover, and spare parts usually dominate cost.
- Which humanoid is best for warehouse work?
- Prefer platforms with paid facility hours and clear payload/end-effector stories. Announced next-gens are not the robots moving goods today.
- Can I buy a home humanoid today?
- Mostly pre-order or prototype. We keep /best/home/ as a stub until independent in-home evidence exists. For the longer view, see House of 47's research brief on home robot timelines (House of 47 is Robotscomparison's parent company).
- Do you sell rankings?
- No. Organic order is not for sale. Sponsored placements are labeled.
- Figure 03 vs Atlas — who wins?
- Neither is a catalog purchase. Split is logistics/partner narrative vs dynamic industrial athletics — compare fields, not a trophy.