Prometheus Brief
Other·Physical AI·Series B·🇺🇸🇬🇧🇨🇭
Summary
Worth a closer lookPrometheus.ai is positioning itself as a platform-level physical‑AI integrator that aims to build a multimodal “artificial general engineer” to automate design and manufacturing; the company is publicly led by co‑CEOs Jeff Bezos and Vik Bajaj with hires such as Kyle Kosic and offices in San Francisco, London and Zurich. It deliberately sits above peers like Intrinsic (robot programming), Covariant (warehouse robotics), AMI Labs (factory optimization), Periodic Labs (materials-discovery labs) and World Labs (supply‑chain AI), claiming an end‑to‑end systems wedge rather than a point solution. The balance-sheet and build signals are extreme: $28.2B raised across three disclosed financings (including a $12B Series B in June 2026) and a reported $41B valuation, plus a 120–150 headcount — clear runway to iterate. The single biggest tension is valuation and ambition versus commercial visibility: no disclosed customers, no stated served markets on the site, and a conservative 3‑year SOM scenario that implies ≈USD 178M versus the reported valuation. For a first meeting, focus on live pilots, IP/tech demos, contractual commitments and the GTM roadmap to see if the engineering scale maps to near‑term revenue pathways.
Bull case
- Extremely deep capital runway: $28.2B disclosed across three rounds (Nov 2025 $6.2B; Apr 2026 $10B; Jun 2026 $12B) and a reported $41B valuation enables multi‑year R&D and deployment programs without immediate sales pressure.
- Founding and technical leadership: public co‑CEOs Jeff Bezos and Vik Bajaj, plus hires such as Kyle Kosic to lead AI infrastructure, and a 120–150 employee base across San Francisco, London and Zurich indicate senior talent and rapid team build.
- Market opportunity and positioning: targets the 2025 Physical AI TAM of USD 81.6B (software & services SAM ≈USD 35.6B), and positions as an end‑to‑end industrial redesign player rather than a single vertical, giving potential access to large, cross‑industry customers.
- Clear momentum signal from consecutive large financings and hiring milestones in the last 12 months, showing investor conviction and operational scale‑up rather than a stealth lab with constrained resources.
Watch-outs
- No public customers, pilots, case studies or third‑party customer reviews were found; the company’s site lists offices but provides no explicit statements about where it sells or who its customers are.
- Valuation vs. disclosed commercial signal: the conservative, traction‑anchored 3‑year SOM scenario (~USD 178M) looks incongruent with a reported $41B valuation and suggests heavy dependence on execution or future capital raises to justify multiples.
- Crowded, specialized competitors occupy adjacent slices (Intrinsic, Covariant, AMI Labs, Periodic Labs, World Labs) — each has productized value in robot retrofits, warehouse automation, factory optimization, materials R&D or supply‑chain execution that could win early enterprise budgets before Prometheus achieves system‑level outcomes.
- Public visibility inconsistencies: external AI assistant descriptions and some public profiles contain factual errors, so core claims (founding, funding, valuation, acquisitions) require primary‑source verification and scrutiny of IP and contractual commitments.
From their website
Stats they publish
Pricing
Not publicly disclosed. The company website does not list its pricing on any of the 6 pages we read.
Target segments
- Trade name
- Prometheus Industries, Inc.
- Registered address
- 2261 Market Street STE 71012, San Francisco, CA 94114
- [email protected]
Founders & Team
3- Jeff BezosFounderCo-chief executive (Co-CEO)Co-founder of Amazon; founder of Blue Origin.
- Vik BajajFounderCo-chief executive (Co-CEO)Chemist and physicist; formerly of Google X; co-founder and CEO of AI-incubator Foresite Labs; adjunct professor of radiology at Stanford University School of Medicine.
- Kyle KosicLead, AI infrastructureCo-founder of xAI; former OpenAI employee.
Prometheus (Project Prometheus) is publicly led by co-founders and co-CEOs Jeff Bezos and Vik Bajaj. Press reports also note hires such as Kyle Kosic to lead AI infrastructure and cite a headcount in the ~120–150 range.
- Several search results reference different companies named 'Prometheus' (Prometheus Group, Prometheus Laboratories, etc.); this extract includes only people explicitly tied to Project Prometheus / Prometheus Industries (the Jeff Bezos/Vik Bajaj AI startup).
- Founders and leadership are taken from press coverage and Wikipedia; an official company team page on prometheus.ai was not provided among the sources.
- Press reports list varying headcounts (Wikipedia: 120; Inc./Ars Technica: ~150).
Market Size
Physical AI market estimates vary widely by publisher; a single, published global figure from Grand View Research places Physical AI at USD 81.6B in 2025 (used here as the TAM). Narrowing to the software & services layer (≈43.6% of the market per SNS Insider) yields a serviceable, global software SAM of ≈USD 35.6B in 2025. A conservative, traction-anchored 3-year SOM scenario (0.5% of that SAM, reflecting Prometheus's disclosed USD 12B Series B and USD 41B valuation) implies ≈USD 178M.
- Grand View Research ↗The global physical AI market was valued at USD 81.6 billion in 2025 and is projected to grow from USD 110.8 billion in 2026 to USD 960.4 billion by 2033, at a CAGR of 36.1% from 2026 to 2033. · Global · 36.1% CAGR
- MarketsandMarkets ↗The market is projected to grow from USD 1.50 billion in 2026 to USD 15.24 billion by 2032, at a CAGR of 47.2% during the forecast period. · Global · 47.2% CAGR
- Acumen Research and Consulting ↗The market was valued at USD 5.02 billion in 2025 and is anticipated to reach around USD 82.79 billion by 2035, indicating a CAGR of 32.8%. · Global · 32.8% CAGR
- SNS Insider (market report) ↗Physical AI Market was valued at USD 5.23 billion in 2025 and is expected to reach USD 87.43 billion by 2035, growing at a CAGR of 32.53% from 2026–2035; hardware dominated with approximately 56.40% of revenues in 2025. · Global · 32.53% CAGR
- Precedence Research (via GlobeNewswire) ↗The global physical AI market size is calculated at USD 5.41 billion in 2025 and is expected to reach around USD 61.19 billion by 2034, growing at a CAGR of 31.26% for the forecasted period. · Global · 31.26% CAGR
- Primary TAM figure usedUSD 81.6 billion global Physical AI market in 2025Grand View Research (see source entry)
- Software & services shareEstimated at ≈43.6% of Physical AI revenues in 2025 (hardware reported ~56.4% of revenues in 2025)SNS Insider market report (hardware 56.4% in 2025) — used to isolate the software/services layer.
- SAM formulaSAM = Global Physical AI TAM (Grand View Research 2025) × software & services share (≈43.6%)Calculation using Grand View Research TAM and SNS Insider component breakdown.
- Company traction (used to ground SOM)Prometheus disclosed USD 12 billion Series B, USD 41 billion valuation, >120 employees (Dec 2025).Prometheus company disclosures (known facts from the company).
- SOM share assumption0.5% of software & services SAM (obtainable over ~3 years); scenario grounded to the company's large Series B and valuation but kept conservative.Judgement anchored to disclosed funding/valuation/size (Prometheus) and typical enterprise adoption timelines for platform-level AI vendors.
- Published Physical AI estimates differ dramatically between research firms (examples in sources) because definitions vary (some reports include hardware, platforms, services; others focus narrowly on software or robotics).
- SAM here is a top-down slice (software & services portion of a full Physical AI TAM) because Prometheus's public pricing/ACV is not available; a bottom-up ACV × customer-count SAM was not possible from disclosed pricing.
- The SOM is a scenario (not a forecast): it applies a conservative obtainable-share (0.5%) anchored to Prometheus's disclosed Series B size, valuation and team scale; this share is judgmental and low-confidence.
- Some source reports are paywalled or use differing base years and forecast windows; reported CAGRs and base-year values are not directly comparable without harmonizing definitions.
Competitors
Top 5Landscape spans robot-software, warehouse automation, factory optimization, materials R&D and supply-chain AI. Prometheus — with $12B Series B and $41B valuation — sits as a platform-level, category-defining industrial-AI integrator pursuing end-to-end manufacturing reinvention.
- Intrinsicintrinsic.ai ↗AI software that makes industrial robots smarter and easier to program; Prometheus targets end-to-end manufacturing redesign and AG-engineer, not robot retrofits.
- Covariantcovariant.ai ↗AI-driven robotic systems for adaptable warehouse automation; Prometheus positions higher, aiming to automate design and manufacturing across industries rather than logistics-only.
- AMI Labsamilabs.xyz ↗Provides ML-driven production optimization and quality-control for factories; Prometheus pursues deeper system-level automation and design, offering broader transformational potential.
- Periodic Labsperiodic.com ↗Builds autonomous labs and AI scientists for materials discovery; Prometheus focuses downstream on industrial design/manufacturing — Periodic wins at experiment-driven materials R&D.
- World Labsworldlabs.ai ↗AI for supply-chain optimization: logistics, inventory, demand forecasting; Prometheus operates upstream on product design/manufacturing, while World Labs wins operational supply-chain execution.
Funding
- initial raise$6.2 billionNovember 18, 2025Jeff BezosTechCrunch ↗
- funding round$10 billionApril 23, 2026JPMorgan ChaseBlackRockInvesting.com ↗
- Series B$12 billionJune 11, 2026JPMorgan ChaseBlackRockGoldman SachsDST GlobalArch Venture PartnersJeff BezosTechCrunch ↗
Prometheus, the physical-AI startup led by Jeff Bezos and Vik Bajaj, has publicly disclosed three large financings: an initial $6.2B at launch in November 2025, a $10B round in April 2026, and a $12B Series B in June 2026. The disclosed rounds sum to $28.2B and the latest reported valuation is $41B.
- Press outlets differ in how they label the early financings (some accounts call the $6.2B a debut/Series A and other coverage treats the $10B as a separate funding event); the round_type values above follow the verbatim phrasing found in the cited articles.
- Investor lists vary by outlet; investors named above are those that specific sources tie to the cited rounds (other named firms appearing in some reports may have participated but were not tied to a specific round in every source).
- A Crunchbase entry surfaced in the research refers to a different "Prometheus/Prometheus Technologies" (an investing/ESG product) at the same domain, indicating name collisions in trackers — that Crunchbase record was not used to attribute the Bezos-led industrial-AI rounds listed here.
- Some outlets previously reported a different total (examples: reports mentioning "more than $18 billion" after earlier rounds); the $28.2B figure above is the sum of the three dated financings itemized in this record and may not capture undisclosed or unreported raises.
Momentum
- FundingJune 11, 2026Prometheus raised $12 billion in a Series B at a $41 billion valuation.TechCrunch ↗
- FundingApril 23, 2026Prometheus finalized a $10 billion funding round (reported April 2026).Investing.com ↗
- FundingNovember 18, 2025Prometheus raised $6.2 billion in an initial funding round.TechCrunch ↗
- MilestoneDec 2025Prometheus had grown to over 120 employees.Company website
Public signals in the last 12 months are dominated by very large, consecutive funding rounds (Nov 2025 → Apr 2026 → Jun 2026) and a hiring milestone in late 2025, indicating accelerating commercial momentum focused on capital and team build-out.
- Provided text includes multiple different organizations named "Prometheus" (open-source Prometheus, Prometheus Group) — only items explicitly tied to the Jeff Bezos / Project Prometheus funding and company facts above were included.
- Coverage supplied here is heavily funding-oriented; operational, partnership, product-launch, or customer announcements are sparse in the provided material.
- Some source snippets reference slightly different dates/wording for the same rounds (e.g., April 21 vs April 23, 2026); dates used reflect the dates stated in the supplied lines.
Customer Sentiment
No public third‑party customer reviews or aggregate ratings for prometheus (prometheus.ai) were found in public review sources; the visible 'Prometheus' references instead concern the 2012 Ridley Scott film and unrelated monitoring/enterprise products named Prometheus.
No third‑party reviews or ratings explicitly for prometheus.ai appear in the public review results examined.
Search results and review snippets found refer to other entities named 'Prometheus' (the 2012 film; the open‑source monitoring project Prometheus; Prometheus Group / Prometheus Platform enterprise software), creating name‑collision risk.
Insufficient direct customer review signal to judge user sentiment for prometheus.ai; common for early‑stage or niche B2B companies.
AI Visibility
We ask ChatGPT WITH web search about the company and show its answers plus the sources it cited — the real AI-search experience. The key signal is share of voice: does the company come up when AI is asked about its space, not just when asked by name.
Models asked ChatGPT · web search
Accuracy of what AI said: Outdated or wrong — Correctly identifies Prometheus's focus on AI for physical/industrial systems, but misstates founders, funding, valuation, and acquisition claims.
Prometheus Industries, Inc., accessible at prometheus.ai, is an artificial intelligence company founded in late 2024 by Jeff Bezos and Vik Bajaj. The company focuses on developing AI models for engineering, manufacturing, and other physical-world applications. Bezos and Bajaj serve as co-CEOs, leading the company's efforts to create AI-powered software designed to optimize engineering workflows, manufacturing processes, and industrial systems. Notably, Prometheus is not building robots itself but instead focuses on the underlying intelligence that can improve productivity and decision-making across physical industries. (citybiz.co)
In June 2026, Prometheus raised $12 billion in a Series B funding round, valuing the company at $41 billion. The company has also acquired General Agents, a startup co-founded by former Google DeepMind researcher Sherjil Ozair, which uses a video-language-action (VLA) model to interpret visual inputs and act on natural language commands—a capability directly applicable to robotics and manufacturing. (citybiz.co)
As of December 2025, Prometheus had hired over 120 employees, including researchers from AI firms such as Meta, OpenAI, and DeepMind. The company is based in San Francisco, with offices in London and Zurich. (en.wikipedia.org)
Prometheus aims to train AI models to understand the physical world by combining visual data, real-world interactions, and complex engineering workflows. This approach is intended to equip autonomous, industrial robots with a deeper, more practical grasp of how things work, thereby improving manufacturing in industries ranging from aviation to drug development. (builtin.com)
The physical AI sector encompasses companies that integrate artificial intelligence with robotics and physical systems. Notable entities and their key contributions include:
- NVIDIA Corporation: Provides AI computing platforms and GPUs that power next-generation robotic intelligence. (global-daily.com)
- Tesla, Inc.: Develops the "Optimus" humanoid robot and advances autonomous vehicle technologies. (global-daily.com)
- Boston Dynamics: Known for humanoid and quadruped robots capable of dynamic movement and real-world adaptability. (global-daily.com)
- Physical Intelligence Inc.: Develops machine learning models intended to control robots and other physical devices. (en.wikipedia.org)
- Figure AI: Specializes in humanoid robots, with products like Figure 03 and Helix, targeting home environments. (aiwiki.ai)
- Agility Robotics: Specializes in bipedal robots designed for logistics and delivery operations. (global-daily.com)
- 1X Technologies: Opened America's first vertically integrated humanoid factory in April 2026. (marketintelo.com)
- Microsoft: Provides critical cloud infrastructure through Azure Robotics and its cross-portfolio investments. (marketintelo.com)
- Google DeepMind: Develops foundation models and robotics research, including Gemini Robotics. (aiwiki.ai)
- Morph: Introduces a shape-shifting soft robotics platform inspired by octopus adaptability. (axios.com)
These companies are at the forefront of integrating AI with physical systems, driving advancements in robotics and automation.
The field of physical AI, which integrates artificial intelligence with physical systems to perform real-world tasks, has seen significant advancements. Here are some notable companies and products in this space:
- Nomagic: Specializes in end-to-end warehouse automation using Physical AI to handle complex tasks with minimal intervention. (nomagic.ai)
- NEMI AI: Offers a comprehensive Physical AI platform for manufacturing, enabling faster design, development, and deployment of physical products. (nemi-ai.com)
- Archetype AI: Develops a Physical AI platform powered by their foundation model, Newton, to deploy intelligent agents across various industries. (archetypeai.io)
- 375ai: Builds a data layer for Physical AI by providing continuous, labeled multimodal data streams from numerous locations, enhancing real-world AI applications. (375.ai)
- Ceva-IP: Provides AI Fabric IP to build and scale Physical AI systems, enabling machines to sense, process, and act in real-time. (ceva-ip.com)
- Wiliot: Offers a Physical AI and IoT platform that enables physical assets to track themselves continuously, sensing various conditions and streaming real-time data without batteries. (wiliot.com)
- Tossom: Develops general-purpose robotics AI foundation models, such as Physical Intelligence (π), enabling robots to perform diverse tasks across different environments. (tossom.com)
- Pickle Robot: Focuses on supply chain automation by deploying robots equipped with generative AI, machine vision, and autonomy to handle tasks like unloading trucks. (picklerobot.com)
- Olive Robotics: Provides embedded Physical AI solutions for sensors and robots, emphasizing interoperability and real-time performance. (olive-robotics.com)
- Dexterity: Enhances robots with Physical AI, enabling them to perform complex tasks safely and efficiently in various applications. (dexterity.ai)
- Universal Robots: Offers a robotics platform for Physical AI, facilitating the development, deployment, and scaling of AI robotics in industrial environments. (universal-robots.com)
Additionally, emerging companies like Morph are developing shape-shifting soft robotics platforms inspired by octopus adaptability, aiming to bring flexible AI applications to the physical world. (axios.com)
These companies are at the forefront of integrating AI with physical systems, driving innovation across various industries.
The assistant recognizes prometheus when asked directly but did not surface it unprompted in category or recommendation lists. Its direct description contains significant factual errors (founders, funding, valuation, and acquisitions) and should not be relied on for diligence without primary-source verification.
- AI answers can vary run-to-run and different prompts or model instances may surface other companies or facts.
- Public web coverage for early-stage or stealth physical-AI firms can be sparse, inconsistent, or conflated with similarly named projects, increasing risk of misinformation.
- Some of the assistant's cited claims reference sources or figures that appear misattributed or unlikely; always confirm founders, funding, and acquisitions from primary filings, press releases, or trusted outlets.
Risks & Red Flags
No material public risk signals found. An absence of public concerns isn’t a clean bill of health — early or private companies may simply not generate coverage.
An absence of public reports is not a clean bill of health: early-stage or privately active companies can have limited public coverage or confidential disputes not visible in public records.
Automated domain-safety/trust-score services have flagged other similarly-named domains (prometheusai.tech and prometheus-ai.online) as high risk; those domains are distinct from prometheus.ai and are not attributable to Prometheus (prometheus.ai). Such algorithmic 'trust score' outputs were excluded from the risk flags because they are not specific incident reports.