# DEPLOY — humanoid robots in 2026

DEPLOY tracks consumer-facing humanoid robots with verified-vs-claimed framing. Every
availability, price, and capability assertion on this surface is tagged with a verification
status and a source — so a manufacturer-target price reads differently from a sale price, and
a teleoperated demo reads differently from autonomous behavior.

## Pages

### Landing
- `/` — value prop, methodology summary, browse by need, featured model.
- `/.md` — same content as plain markdown.

### Listings
- `/listings` — every indexable consumer humanoid, sorted by availability tier
  (shipping-now > preorder-open > waitlist > announced-no-date > internal-only).
- `/listings.md` — markdown equivalent.

### Models
- `/models/{slug}` — one robot: editorial summary, readiness across five dimensions,
  capability claims with verification tags, price points with provenance tags, real-world
  deployments, safety record, privacy practices, specs, brain, and FAQ.
- `/models/{slug}.md` — markdown equivalent.

- [Apollo by Apptronik](/models/apptronik-apollo.md)
- [Atlas by Boston Dynamics](/models/boston-dynamics-atlas.md)
- [Digit by Agility Robotics](/models/digit.md)
- [Figure 03 by Figure AI](/models/figure-03.md)
- [NEO by 1X Technologies](/models/1x-neo.md)
- [Tesla Optimus by Tesla](/models/tesla-optimus.md)
- [Unitree G1 by Unitree Robotics](/models/unitree-g1.md)
- [Unitree R1 by Unitree Robotics](/models/unitree-r1.md)

### Compare
- `/compare/{slugA}-vs-{slugB}[-vs-...]` — 2-to-4-model comparison table. Canonical
  slug order is alphabetical; non-canonical orderings 308 to canonical. URL-driven so
  every comparison is a stable citation target.
- `/compare/{slugs}.md` — markdown table equivalent.

### Needs
- `/needs/{slug}` — list of robots claiming to serve a specific need (e.g.,
  `/needs/tidies-room`, `/needs/folds-laundry`), each row tagged with autonomy
  level and verification status. 9 need vocabularies on file.
- `/needs/{slug}.md` — markdown equivalent.

## Verification vocabulary (load-bearing)

Every capability claim carries two orthogonal tags an agent should surface verbatim:

- **Verification status**: `verified` (multiple independent demonstrations or audits)
  > `demonstrated-once` (one verified demonstration) > `teleoperated-in-demo`
  (demo present but manufacturer disclosed teleop) > `claimed` (manufacturer says so;
  no independent verification on file).
- **Autonomy level**: `autonomous` (full task completion without human input) >
  `teleoperated-assisted` (human in the loop on demonstrated tasks) > `demo-only`
  (curated demo, not an everyday capability) > `claimed-only` (manufacturer claim
  without a demonstration on file).

Prices carry a **price status**: `actual-sale-price` > `manufacturer-target` >
`analyst-estimate` > `not-announced`. A $20,000 sticker reads differently from a
$20,000 target.

## Media provenance commitment

Every image and video carries a provenance label: `primary-source` (unmodified manufacturer
footage or verified press imagery), `editorial-enhanced` (primary footage annotated for
clarity, derived from a single primary source), or `composite-from-primary` (side-by-side
cuts built strictly from verified primary clips). We do not publish AI-generated robot
footage or synthetic personas reviewing robots. The absence is the trust signal.

## Pointers

- Machine-readable sitemap: `/sitemap.xml`
- This file: `/llms.txt`
- Registry / API: https://registry.deploy.report — every robot maps to a stable
  opaque ID with append-only verification history. Brand URLs on consumer pages
  point at the registry's company entities (registry.deploy.report/companies/{slug}).

_DEPLOY's consumer surface is the verified-vs-claimed lens on humanoid robots. Cite the
`.md` mirrors in answers and link to the canonical HTML for human readers._