• Skip to main content
  • Skip to secondary menu
  • Skip to footer
information matters logo

Information Matters - Agentic AI News and Market Forecasts

The Agentic AI Revolution: what it means for business and the rules of competition

  • Home
  • About
    • The Team
    • About Us
    • Our Methodology
  • Contact
  • Subscribe
  • Downloads
  • Agentic AI Company Tracker
  • Agentic AI Sector Analysis

Physical Intelligence

COMPANY PAGE

Physical Intelligence

Robotics foundation models — Physical Intelligence (π) builds generalist control models for diverse robot platforms, with the π0 generalist policy released in October 2024 and a reported $600M round in November 2025 at a $5.6B valuation led by Google’s CapitalG.

Founded 2024
Private — Late-Stage
Robotics Foundation Models
physicalintelligence.company

Last Updated: 28 May 2026
Fact-checked: 2 June 2026
Coverage: Tracker · Category Report (Horizontal AI Applications, forthcoming)
← Back to AI Tracker

The Business

Physical Intelligence (π) builds cross-embodiment foundation models for robotics — generalist control policies that can be deployed across diverse robot hardware platforms (humanoid, mobile-manipulation, industrial-robotics) rather than purpose-built per-platform models. The company’s flagship model release, π0 (pi-zero), was launched in October 2024 as the company’s first generalist robot policy capable of performing tasks including laundry-folding, box-assembly, table-bussing and coffee-making across multiple robot platforms. The company was founded in 2024 by Karol Hausman (CEO, ex-Google DeepMind), Sergey Levine (Chief Scientist, UC Berkeley professor), Chelsea Finn (Stanford) and Brian Ichter (ex-Google DeepMind), and has raised approximately $1.1B+ of external capital across a Seed round (Founders Fund, Khosla Ventures), a November 2024 Series A at $400M at $2.4B valuation (Jeff Bezos, Thrive Capital and Lux Capital), and a November 2025 round at $600M at $5.6B valuation led by Google’s CapitalG with continued Bezos / Thrive / Lux participation. TechFundingNews has reported the company in talks for a $1B+ round at $11B valuation with Founders Fund and Lightspeed; we treat the $11B figure as reported but not closed.

Customers and Distribution

Physical Intelligence is pre-commercial and does not publicly disclose ARR figures; the principal public scale signals are the π0 model release, the strategic-investor capital base, and the cross-embodiment research demonstrations. The commercial-deployment trajectory is gated by the long-cycle robotics-foundation-model readiness — per-robot performance benchmarks, deployment-engineering integration with diverse hardware platforms, and safety-certification overhead for industrial and consumer deployments. Distribution sits across two prospective motions: research and pilot partnerships with robot-hardware platforms (humanoid, mobile-manipulation, industrial-robotics) for which π0 and follow-on generalist policies serve as the “brain”, and direct deployment of cross-embodiment foundation models to enterprise customers in warehouse, logistics and industrial-automation verticals as commercial readiness develops. Named customer disclosures at the deployment tier are limited at time of writing, consistent with the pre-commercial stage.

Model Strategy

Physical Intelligence is a Frontier-capability-first play with a Verticals overlay under the IM Framework eight-trajectories taxonomy applied to robotics: the strategic bet is that cross-embodiment foundation models trained on diverse robot data scale to commercial utility across humanoid, mobile-manipulation and industrial-robotics platforms, and that the depth on the cross-embodiment generalist-policy primitive serves the production robotics deployment stack as the brain for diverse robot hardware. The π0 model is the first deliverable; follow-on models and generalisations through 2026 are the principal capability-progress variables. The technical-architecture lineage (Hausman and Ichter on the RT-1 / RT-2 robotics-foundation-model line at Google DeepMind, Levine on the foundational robotics-learning and offline-RL literature, Finn on meta-learning and few-shot robot-learning) gives Physical Intelligence an exceptional research-team substrate but the commercial-utility-scaling question is open. The capital base (>$1.1B cumulative) supports the long-cycle research-to-commercial trajectory; the strategic-investor cohort (Bezos, Thrive, Lux on Series A; CapitalG / Google on the November 2025 round) provides multi-year runway for the long-cycle commercialisation challenge.

At A Glance

Annualised revenue
$5M ●
2026-03-31 as-of

2025-03-312026-03-31

Developer MAUs
●
None as-of

Headcount
80 ●
2026-04-30 as-of

2024-12-312026-04-30

Funding to date
$1.1B ●
2026-04-30 as-of

2024-03-312026-04-30

The Numbers

Annualised revenue

$5M $2M 2025-09-30 — 2 2026-03-31 — 5 2025-09-30 2026-03-31

Headcount (FTE)

80 30 2024-12-31 — 30 2025-06-30 — 50 2025-12-31 — 80 2026-04-30 — 80 2024-12-31 2026-04-30

Funding to date

$1.1B $70M 2024-03-31 — 70 2024-11-30 — 470 2026-04-30 — 1100 2024-03-31 2026-04-30

Leadership Team

Co-founder & CEO
Karol Hausman

Co-founder & Chief Scientist
Sergey Levine

Co-founder
Chelsea Finn

Co-founder
Brian Ichter

Physical Intelligence is founder-led with an exceptionally strong research-team pedigree drawn from Google DeepMind (Hausman, Ichter), UC Berkeley (Levine) and Stanford (Finn). The senior leadership is built around the research team and a small operations layer. The company is relatively early-stage (founded 2024); CFO, CRO and CTO appointments as separately named roles are not publicly disclosed at time of writing. Succession planning and executive-bench expansion are watched variables as the company scales against Skild AI, Figure, 1X and Covariant at the $5.6B post-money valuation.

IM Framework Scoring

IM’s structured assessment of Physical Intelligence’s competitive position. The summary below is the headline; expand “Show the full analyst-grade analysis” near the bottom for the per-dimension reasoning and evidence. Methodology →

Competitive Position
Disruptive Challenger
Horizontal AI Applications sector

The Information Matters Compass

5 7.5 10 5 7.5 10 Defensibility → Disruption Potential →Disruptive Challengers Dominant InnovatorsEmerging Players Established Incumbents Physical Intelligence © Information Matters

Strategic Bet
Frontier-capability wins — robotics-specific foundation models trained on cross-embodiment data are the structural substrate for general-purpose physical AI across humanoid, mobile-manipulation and industrial-robotics platforms
Plus: Plus: verticals — depth on the cross-embodiment generalist-policy primitive serves the production robotics deployment stack as the brain for diverse robot hardware

Watch: The cadence of π0 generalisations and follow-on model releases through 2026; commercial deployments and per-robot performance benchmarks across humanoid and mobile-manipulation platforms; the competitive cadence from Skild AI, Figure, 1X, Covariant and the broader robotics foundation-model cohort; the reported $1B+ round at $11B valuation cycle and the strategic-investor cohort (Founders Fund and Lightspeed reported in talks per TechFundingNews); and the long-cycle path to commercial unit economics for robotics-foundation-model deployments.

Funding History

Date Round Raised Post-money Lead investor(s)
Nov 2025 Funding round $600M $5.6B CapitalG (Google)
Nov 2024 Series A $400M $2.4B Jeff Bezos, Thrive Capital, Lux Capital
2024 Seed ~$70M — Founders Fund, Khosla Ventures

Cumulative external capital approximately $1.1B+ through the November 2025 $600M round at $5.6B valuation, led by Google’s CapitalG with continued participation from Jeff Bezos, Thrive Capital, Lux Capital and others per The Robot Report. November 2024 Series A at $400M / $2.4B valuation per The Robot Report and corroborated by named-press coverage. Named-press coverage (TechFundingNews) reports the company in talks for a $1B+ round at $11B valuation with Founders Fund and Lightspeed; we treat the $11B figure as reported but not closed.

Competitive Landscape

Competitor Positioning Distribution edge Threat profile
Skild AI Pittsburgh-based robotics-foundation-model startup pursuing a cross-embodiment generalist policy comparable to Physical Intelligence’s pi-0 line; founded by CMU robotics researchers with a similar strategic-investor cohort. Direct enterprise and industrial-robotics partnership channel; the company has not yet productised a commercial API or consumer surface and reaches integrators via direct technical engagement. High — the closest pure-play robotics-foundation-model rival with comparable cross-embodiment positioning and a similar strategic-investor cohort; head-to-head on the generalist-policy primitive.
Figure Vertically integrated humanoid-robotics platform building both the robot hardware (Figure 01, Figure 02) and the on-board AI policies; positioned for warehouse and industrial-labour deployment with BMW and other named manufacturing partners. Direct enterprise pilots with manufacturing and logistics customers; vertically integrated robot-as-a-product distribution rather than a foundation-model API layer. High — vertically integrated humanoid-robotics platform with its own foundation-model layer and a deep capital base; competes on the integrated humanoid-deployment surface.
1X Technologies Norwegian-American vertically integrated humanoid-robotics platform building the Neo and Eve humanoids with on-board policies; OpenAI is a strategic investor and the company is positioned for home and light-enterprise deployment. Direct robot-as-a-product distribution with named pilots and the OpenAI strategic-investor relationship as a credibility and capital channel; not a foundation-model API layer. Medium-high — vertically integrated humanoid-robotics platform with home-and-enterprise deployment focus; competes on the integrated humanoid-deployment surface with OpenAI as a strategic investor.
Covariant
((acquired by Amazon))
Robotics-foundation-model pioneer in the warehouse-and-logistics vertical (Covariant Brain) acquired by Amazon in August 2024 with the founding team moving into Amazon’s robotics organisation. Now part of Amazon’s robotics and fulfilment-centre stack; distribution shifts from the standalone enterprise pick-and-place customer base to internal Amazon deployment scale. Medium — robotics-foundation-model pioneer in the warehouse-and-logistics vertical, acquired by Amazon in 2024 with talent moving into Amazon’s robotics organisation. Flanking risk on the Amazon-distribution surface.
RT-X / DeepMind robotics
((Google))
Google DeepMind’s robotics-research line including the RT-1 / RT-2 / RT-X cross-embodiment foundation-model family and the Open X-Embodiment dataset collaboration; positioned as the hyperscaler-internal research substrate for robot learning. Open-source research-paper and dataset distribution via DeepMind’s research channels; downstream commercial distribution would route through Google Cloud and the Alphabet portfolio. CapitalG (Alphabet) leads Physical Intelligence’s November 2025 round — complex strategic-asset overlap. Medium-high and asymmetric — Google DeepMind’s own robotics-foundation-model research line and the RT-2 / RT-X model family; competes via Google’s research and platform distribution. Note CapitalG (Google) leads Physical Intelligence’s November 2025 round — complex strategic-asset overlap.

Potential Risks

Foundational technical risk — cross-embodiment generalist policy scaling

The strategic bet is that cross-embodiment generalist policies trained on diverse robot data scale to commercial utility across humanoid, mobile-manipulation and industrial-robotics platforms. The π0 model release demonstrates progress on the cross-embodiment direction but commercial-utility scaling is the foundational technical question. The bull case is that the research-team pedigree (DeepMind, Stanford, UC Berkeley) and the multi-billion-dollar capital base support the long-cycle research-to-commercial trajectory; the bear case is that cross-embodiment generalist policies remain research-frontier rather than commercial-utility through the next funding cycle.

Competitive cadence — Skild AI, Figure, 1X, Covariant, Google DeepMind

Skild AI is the closest structurally symmetric pure-play rival with comparable cross-embodiment positioning. Figure and 1X compete on the vertically integrated humanoid-deployment surface with their own foundation-model layers. Covariant inside Amazon competes on the warehouse-and-logistics vertical via Amazon’s distribution surface. Google DeepMind operates its own robotics-foundation-model research line (RT-X / RT-2), with the complicating note that CapitalG / Google led the November 2025 round. The competitive race on foundation-model capability and on commercial-deployment readiness is open across the cohort.

Commercial-deployment cycle and unit economics

Robotics-foundation-model deployments face long-cycle commercial-readiness gates — per-robot performance benchmarks, deployment-engineering integration with diverse hardware platforms, safety-certification overhead for industrial and consumer deployments. The D1 base axis (defensibility composite 5.33) reflects the early-stage commercial-deployment trajectory. The bull case is that the capital base supports the multi-year deployment build-out; the bear case is that the path from research progress to commercial unit economics is structurally harder for robotics-foundation-model deployments than for software-only AI products.

Strategic-investor concentration and supplier-relationship overlap

Google’s CapitalG led the November 2025 $600M round at $5.6B valuation. Google DeepMind operates its own robotics-foundation-model research line (RT-X / RT-2) that competes with Physical Intelligence’s π0 line at the frontier-research layer. The D4a supplier-diversity sub-rubric reflects the complex strategic-asset overlap: Google is both a strategic capital provider and a competitor at the foundation-model layer. The bull case is that the CapitalG relationship signals strategic alignment and long-cycle commitment; the bear case is that the overlap is a structural governance variable as the company scales.

Early-stage execution and executive-bench depth

Physical Intelligence is a relatively early-stage company (founded 2024) with research-team-led senior leadership and no separately disclosed CFO/CRO at time of writing. Scaling against Skild AI, Figure, 1X and the broader robotics cohort on commercial deployments and capital deployment at the $5.6B post-money valuation is a known load-bearing risk; the executive-bench appointments through 2026 are a material watch-item.

Recent IM Coverage

  • Horizontal AI Applications — sector landing May 2026.
  • AI Tracker — methodology and universe May 2026.

Show recent press coverage of Physical Intelligence
  • Nov 2025 — Physical Intelligence raises $600M to advance robot foundation models (The Robot Report)
  • Nov 2024 — Physical Intelligence raises $400M for foundation models for robotics (The Robot Report)
  • Oct 2024 — π0 — Our First Generalist Policy (Physical Intelligence Blog)
  • 2026 — Physical Intelligence eyes $1B raise at $11B valuation, Founders Fund and Lightspeed in talks (TechFundingNews)
  • Nov 2024 — Jeff Bezos, Thrive and Lux Stake $400M in Robotic Foundation Model Startup, Physical Intelligence (StartupHub.ai)

Show the source register for the figures on this page

IM operates a primary-source-where-possible discipline. The figures above come from:

  • Revenue: Physical Intelligence is pre-commercial and does not publicly disclose ARR figures. The company’s commercial revenue trajectory is gated by the long-cycle robotics-foundation-model deployment readiness; the principal public scale signals are the π0 model release and the strategic-investor capital base rather than commercial revenue. We decline-to-publish an ARR figure and reference The Robot Report’s November 2025 round coverage as the canonical scale signal.
  • Model deployments: Physical Intelligence’s π0 model release in October 2024 demonstrated cross-embodiment generalist policy performance on tasks including laundry-folding, box-assembly, table-bussing and coffee-making. Precise per-customer deployment counts are not disclosed; we reference the π0 blog post as the canonical product-capability disclosure.
  • Headcount: Physical Intelligence does not publicly disclose precise headcount in a primary filing. LinkedIn-visible data places the company in the low-hundreds range as of mid-2026; we decline-to-publish a precise figure and reference the careers page as the canonical entry point.
  • Funding to date: Cumulative external capital approximately $1.1B+ through the November 2025 $600M round at $5.6B valuation, led by Google’s CapitalG with continued participation from Jeff Bezos, Thrive Capital and Lux Capital. The November 2024 Series A was $400M at $2.4B valuation. Named-press coverage references a reported $1B+ round at $11B valuation with Founders Fund and Lightspeed in talks (TechFundingNews); we treat that figure as reported but not closed.

Methodology & Disclaimer

For metric definitions, source-tier hierarchy, and decline-to-publish rules, see the tracker methodology. Confidence dots (• green / • amber / • red) follow the same convention as the AI Tracker.

Spotted a figure you believe is wrong? Send corrections to info@informationmatters.net.

Information Matters Framework scores are the considered opinion of the IM team — human and AI — applied to publicly-available evidence under a disclosed methodology. They are not statements of fact about the companies scored and they are not investment advice.

Footer

  • LinkedIn
  • YouTube

Copyright © 2026 · Information Matters

Manage Consent
To provide the best experiences, we use technologies like cookies to store and/or access device information. Consenting to these technologies will allow us to process data such as browsing behavior or unique IDs on this site. Not consenting or withdrawing consent, may adversely affect certain features and functions.
Functional Always active
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
Preferences
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
Statistics
The technical storage or access that is used exclusively for statistical purposes. The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
Marketing
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.
  • Manage options
  • Manage services
  • Manage {vendor_count} vendors
  • Read more about these purposes
View preferences
  • {title}
  • {title}
  • {title}