AI agents · OpenClaw · self-hosting · automation

Quick Answer

Boston Dynamics vs Figure vs Agility vs 1X (Humanoid Robots July 2026)

Published:

Boston Dynamics vs Figure vs Agility vs 1X: Humanoid Landscape (July 2026)

Hyundai took 100% of Boston Dynamics on July 15, 2026, and reshuffled the humanoid robotics race. With Figure AI valued at ~$40B, Agility deploying Digit at Amazon, 1X shipping Neo, and Tesla Optimus in factories, the question for enterprises and investors is: which humanoid actually wins, and in which market?

Here is the July 2026 head-to-head.

Last verified: July 16, 2026

Quick Comparison

Boston Dynamics AtlasFigure 03Agility Digit v51X NeoTesla Optimus Gen 3
Owner / BackerHyundai (100%, July 15, 2026)Independent, BMW/OpenAI-backedIndependent, Amazon/PlaygroundOpenAI + Tiger GlobalTesla
Form factorElectric humanoidElectric humanoidBipedal, purpose-built for logisticsSlim humanoidHumanoid
Primary use caseIndustrial (Hyundai/Kia lines)Automotive manufacturingWarehouse logisticsHome + light industryTesla factories → external
Real commercial deploymentsRamping (Hyundai)BMW pilotAmazon, GXO LogisticsLimitedTesla only
Locomotion sophisticationHighestHighPurpose-optimizedSolidImproving fast
Dexterous manipulationVery goodVery good (Helix VLA)Good (logistics-tuned)SolidImproving
AI backboneNVIDIA Cosmos + Google/Anthropic likelyHelix (in-house) + OpenAIIn-house + NVIDIA computeGPT-basedTesla FSD lineage
Public listing status2027-2028 IPO candidatePossible 2027Possible 2027Not near-termPublic via Tesla (TSLA)
Rough valuationIncluded in Hyundai~$40B (early 2026 round)Multi-billion, privateMulti-billion, privatePart of Tesla
Unit economics viable?Path clear (Hyundai captive)BMW pilot data mixedBest deployed unit economics todayNot yetTesla claims viable

Boston Dynamics Atlas — The Newly Focused Player

After 13 years of ownership churn (Google → SoftBank → Hyundai partial → Hyundai full), Boston Dynamics finally has a single decisive owner with capital and a captive customer.

Strengths:

  • Deepest locomotion IP — 15+ years of R&D, industry-best whole-body dynamics
  • Captive first customer — Hyundai/Kia manufacturing lines are meaningful volume opportunity
  • NVIDIA Cosmos alignment — Hyundai joined the Cosmos Alliance July 15, 2026
  • Unified capital — no more SoftBank governance drag
  • Path to IPO — likely 2027-2028 window

Weaknesses:

  • Later to real commercial deployment than Agility or Figure
  • Industrial focus — no near-term consumer story
  • Depends on Hyundai’s willingness to be the reference customer (usually good, but not automatic)

Best for: heavy industrial buyers who value locomotion maturity and want a supplier backed by a major OEM.

Figure 03 — The BMW-Backed Momentum Story

Figure AI is the most-hyped humanoid company. Figure 03 pairs improved dexterity with the Helix vision-language-action model — Figure’s in-house VLA that competes with Google’s Gemini Robotics and openai-partnered stacks.

Strengths:

  • Highest valuation of independent humanoid makers (~$40B early 2026 round)
  • Real BMW pilot — automotive manufacturing use case, high production-line stakes
  • Vertical integration — Helix model + hardware + deployment all in-house
  • Strong OpenAI partnership — access to frontier reasoning for higher-level task planning
  • Fast iteration cadence — Figure 01 → 02 → 03 in about 24 months

Weaknesses:

  • BMW pilot data is mixed — reporting suggests slower deployment than public narrative implies
  • High burn rate — hardware + custom model training is capital-intensive
  • Independent risk — no captive OEM parent means IPO or M&A pressure in 2027

Best for: automotive and manufacturing buyers, investors seeking pure-play humanoid exposure.

Agility Digit — The Deployed-in-Anger Winner

Digit is the least Hollywood-looking humanoid and the most deployed. It is bipedal, warehouse-purpose-built, and boring — in the best possible way.

Strengths:

  • Actual deployed unit economics at Amazon fulfillment centers and GXO Logistics
  • Purpose-built for logistics — not trying to do everything, just moving totes
  • Amazon as backer + customer — best-case combination
  • Mature product — Digit v5 is the fifth major iteration
  • NVIDIA-integrated — heavy compute partnership

Weaknesses:

  • Narrow use case — logistics is a large market but not universal
  • Not humanoid-humanoid — hands and face are utilitarian, not a general-purpose worker
  • Less multi-modal AI — Digit is more classical robotics than VLA-driven

Best for: warehouse and logistics operators who need working robots now, not in 2028.

1X Neo — The Home Robot Play

1X’s Neo is the humanoid closest to consumer / home use. Soft covers, safer around humans, slimmer form factor.

Strengths:

  • OpenAI backing — direct access to GPT-based agent reasoning
  • Consumer story — the only credible near-term home humanoid
  • Safety-first design — soft outer skin, lower peak forces
  • Distinct market — not competing directly with Figure/Boston Dynamics in industrial

Weaknesses:

  • Consumer humanoid TAM is aspirational — nobody has proven the price point works
  • Deployment volume is small
  • Compute-per-Neo cost is high — home units cannot amortize compute the way industrial fleets can

Best for: OpenAI ecosystem, home / light industry pilots, researchers.

Tesla Optimus Gen 3 — The Wild Card

Tesla Optimus is deployed at Tesla factories, and the roadmap is aggressive: consumer availability at $20-$30K “eventually.”

Strengths:

  • Tesla manufacturing captive customer — no other maker has this scale
  • FSD-lineage AI — decades of neural net training compute plus Dojo
  • Public equity access — buy TSLA, get Optimus exposure
  • Vertical integration — motors, batteries, compute all Tesla-owned

Weaknesses:

  • External customer traction unverified
  • Elon Musk timeline risk — public promises frequently slip
  • Not clear whether Optimus is a product or a Tesla-internal robot

Best for: Tesla shareholders, believers in Musk timelines.

Chinese Humanoids — The Elephant in the Room

WAIC 2026 (Shanghai, July 17-20) is expected to showcase major Chinese humanoid progress. Key names:

  • Unitree G1 — commercially available at ~$16K, aggressively priced
  • Unitree H1 — larger, industrial variant
  • UBTech Walker S2 — automotive manufacturing pilots at multiple Chinese OEMs
  • Fourier Intelligence GR-1 — research + light industrial
  • Xpeng Iron — automotive maker’s own humanoid

The Chinese cost structure will pressure Western industrial humanoids materially. Expect Western humanoids to compete on locomotion sophistication and enterprise integration; Chinese humanoids to win on volume industrial pricing outside export-restricted markets.

Decision Matrix

BuyerBest Choice
Amazon-scale warehouse operatorAgility Digit (proven at Amazon)
Automotive manufacturerFigure 03 (BMW pilot) or Boston Dynamics Atlas (Hyundai/Kia)
Heavy industrial (steel, chemical, energy)Boston Dynamics Atlas (locomotion maturity matters)
Home / consumer research1X Neo (only real option)
Tesla-adjacent operationsTesla Optimus
Chinese-domestic manufacturerUnitree, UBTech, Xpeng Iron (cost + local support)
Pure investor exposureTSLA public or Figure private (waiting for IPO)
Startup building humanoid software / appsSupport all APIs; pick 2 hardware partners

Where This Goes in 12 Months

  • Boston Dynamics IPO filing — likely late 2027 given typical 12-18 month post-restructuring timelines
  • Figure AI Series D — probable, at or above $50B valuation
  • Agility Digit at 10K+ units deployed — plausible by end-2027
  • NVIDIA Cosmos as de facto physical AI standard — likely
  • Chinese humanoid volume ramp — WAIC 2026 announcements will set the tone

The humanoid market is bifurcating: US/Western makers dominating high-locomotion / enterprise-integration segments, Chinese makers dominating volume industrial pricing. Boston Dynamics under Hyundai is the biggest bet on the Western side. Agility is the biggest deployment story. Figure is the biggest brand narrative.

The Frame

Humanoid robotics in July 2026 is where autonomous vehicles were in 2018 — real deployments happening, most of the market still ahead, valuations aggressive, timelines slippery. The Hyundai/Boston Dynamics deal is the strongest signal yet that this becomes a public-markets story within 24 months.

Do not confuse capability demos with commercial reality. Ask about actual unit deployments, actual maintenance costs, actual hours-of-autonomous-work-per-day. Everyone shows highlight reels. Only Agility and Figure have real production data. Boston Dynamics is about to.

Sources