Startup Funding Watch: Manus’s $500M+ Round, Oratomic’s $475M and Venture Capital’s AI Concentration
Startup Funding Watch · October 9, 2026
Manus Raises $500M+ After Its Meta Exit, Oratomic Lands $475M for Quantum, and Cybersecurity Startups Pull in $5.26B in Q3 — as AI Captures the Majority of Global Venture Capital
Venture Capital Is Concentrating in AI and Infrastructure — and the Headline Totals Obscure It
The quarter's global funding numbers look extraordinary. Venture-backed companies raised approximately $159 billion in Q3 2026, bringing the first three quarters to roughly $679 billion, per Crunchbase. The quarter also posted a record number of billion-dollar funding rounds. But the totals conceal a market that is becoming far more concentrated. AI infrastructure, data centers, semiconductors, robotics and energy are absorbing the overwhelming share of capital — and the companies outside those categories are competing for a shrinking slice of investor attention.
This week's individual rounds reflect that concentration. Manus raised more than $500 million after unwinding its acquisition by Meta, in a round led by Boyu Capital and IDG Capital with Tencent, HSG and ZhenFund participating, per Reuters and TechCrunch — a notable case of an AI-agent startup rebuilding independence after a major acquisition collapsed. Oratomic raised $475 million at a reported $5.4 billion valuation for fault-tolerant quantum computing, with investors including ARCH Venture Partners, Spark Capital, Khosla Ventures, Index Ventures, General Catalyst and Bezos Expeditions, per The Wall Street Journal. And Nous Research raised $90 million at a reported $1.5 billion valuation for open-source AI agents, with Nvidia, Microsoft's M12, Samsung and 1789 Capital participating, per The Wall Street Journal.
Two adjacent categories showed the same pattern. Cybersecurity startup funding reached $5.26 billion in Q3 2026, a reported 78% year-over-year increase, with $12.7 billion raised through September, per The Wall Street Journal. And in defense technology, Gallatin AI raised $50 million in a Series A led by 8VC and Silent Ventures for military logistics software, per Axios — a signal that defense investment is broadening beyond drones and autonomous weapons into the software systems that sustain operations. The unifying question: where is venture capital actually flowing, and which startups are turning funding into defensible technology businesses?
Funding at a Glance
| Company | Amount | Category | Valuation |
|---|---|---|---|
| Manus | $500M+ | AI Agents | Undisclosed |
| Oratomic | $475M Series B | Quantum Computing | ~$5.4B |
| Nous Research | $90M Series B | Open-Source AI | ~$1.5B |
| Gallatin AI | $50M Series A | Defense Software | Undisclosed |
| Q3 cybersecurity total | $5.26B (quarter) | AI Security / Identity | N/A |
| Q3 global VC total | $159B (quarter) | Cross-sector | N/A |
Signal summary: The largest rounds this week went to AI agents, quantum computing and open-source AI — all capital-intensive, technically differentiated categories. The market is rewarding depth over breadth, and the companies outside the AI and infrastructure categories face a materially harder fundraising environment.
Biggest Funding Deals
Manus Raises More Than $500 Million as It Rebuilds Independence After Meta
Company: Manus | Amount: More than $500 million | Lead investors: Boyu Capital and IDG Capital | Participating: Tencent, HSG, ZhenFund | What it does: AI-agent software. The round is the company's first major financing since unwinding its acquisition by Meta.
Why investors are backing it: This is the lead story because it raises a question that has no clear precedent: can an AI-agent startup maintain its growth and strategic independence after a major acquisition falls apart? Manus was positioned to be absorbed by Meta. That path is now closed, and the company is rebuilding as an independent operator with substantial fresh capital.
The strategic significance: Acquisitions that unwind are unusual and often disruptive — teams scatter, momentum stalls, and competitors exploit the uncertainty. Manus raising $500M+ shortly after the unwinding suggests investors believe the underlying technology and market position remain strong, and that independence may in fact be preferable in a market where strategic optionality is valuable.
The investor composition: The participation of Tencent, HSG and ZhenFund — combined with lead investors Boyu Capital and IDG Capital — indicates significant Chinese institutional backing for an AI-agent company at a moment when the global AI market is bifurcating along geopolitical lines.
What the money enables: The company plans to scale its AI-agent business and expand international hiring. Both are capital-intensive and signal an ambition to compete beyond a single market. Path to IPO: Reports indicate the company is positioning for a potential future IPO, which would provide an exit path that the Meta acquisition would have precluded.
Editorial question: Does independence from a large acquirer strengthen Manus's competitive position, or does it leave the company without the distribution and infrastructure advantages that a major parent could provide? Connect to: AI Watch · Company Intelligence · IPO Watch
Oratomic Secures $475 Million as Quantum Computing Attracts More Capital
Company: Oratomic | Stage: Series B | Amount: $475 million | Valuation: ~$5.4 billion (reported) | Investors: ARCH Venture Partners, Spark Capital, Khosla Ventures, Index Ventures, General Catalyst, Bezos Expeditions | What it does: Fault-tolerant quantum computing, with a focus on error correction — the technical problem that has to be solved before quantum computers can perform useful work at scale.
Why investors are backing it: The investor list is unusually strong for a quantum computing company, spanning deep-tech specialists (ARCH), generalist venture firms (Index, General Catalyst), and high-profile individual investors (Bezos Expeditions). That composition suggests conviction that quantum computing is approaching a genuine commercial inflection rather than remaining a perpetual research project.
The technical milestone: Fault-tolerant quantum computing requires error rates low enough that logical qubits can be maintained reliably. That is the threshold at which quantum systems become useful for problems classical computers cannot solve — chemistry simulation, optimisation, and certain cryptographic applications. Oratomic is explicitly targeting that threshold.
What the money enables: The round funds the engineering work required to turn research results into a manufacturable system. That is a materially different task from demonstrating a physics result in a laboratory — it requires repeatability, yield, and integration at scale.
The comparison to AI infrastructure: Quantum computing is attracting capital at a scale that resembles early AI infrastructure investment, but the timelines are longer, and the commercial applications are less defined. That does not make the investment wrong — it makes it a different risk profile.
Editorial question: Does Oratomic's valuation reflect near-term commercial potential, or does it price in a quantum computing market that is still years from materialising at scale? Editorial note: Large funding rounds accelerate technical development, but valuation is not proof of commercial readiness. Milestone progress should be tracked rather than assumed. Connect to: Semiconductor Watch · AI Infrastructure Special Report · VC Intelligence
Nous Research Raises $90 Million as Investors Back Open-Source AI Agents
Company: Nous Research | Stage: Series B | Amount: $90 million | Valuation: ~$1.5 billion (reported) | Lead investor: Robot Ventures | Participating: Nvidia, Microsoft's M12, Samsung, 1789 Capital | What it does: Develops the Hermes open-source AI assistant. The company is positioned as an alternative to closed, proprietary AI platforms, focused on customizable agents with greater user control.
Why investors are backing it: The investor composition is the most informative detail. Nvidia, Microsoft's M12, and Samsung — three companies with direct stakes in the AI hardware and platform ecosystem — are all participating. That suggests they view open-source AI agents not as a competitive threat to their core businesses but as a complement that expands total AI adoption.
The open-source thesis: Enterprises increasingly want AI systems they can customize, run on their own infrastructure, and audit. Closed platforms offer capability but limit control. Open-source models offer control but require more internal capability to deploy. Nous Research's positioning sits in the middle — open weights with a supported product.
The economics question: Open-source AI businesses face a structural challenge: the models themselves are freely available, so the value must come from support, integration, hosting, or enterprise features. That is a harder business model to defend than a proprietary platform, but it also scales differently — open-source adoption can grow without per-seat pricing.
The competitive context: Investor interest in open-source AI is rising as enterprises express frustration with the cost and lock-in of proprietary platforms. But adoption will depend on reliability, security, support, and the cost of running models — not on licensing philosophy alone.
Editorial question: Do open-source AI agents become the enterprise default for organizations that want control, or do they remain a specialized choice for technically sophisticated buyers? Connect to: AI Watch · Enterprise AI Intelligence · Semiconductor Watch
Gallatin AI Raises $50 Million to Modernize Military Logistics
Company: Gallatin AI | Stage: Series A | Amount: $50 million | Lead investors: 8VC and Silent Ventures | What it does: Develops the Navigator product, focused on military logistics and sustainment software.
Why investors are backing it: Defense technology investment has broadened well beyond drones and autonomous weapons. Gallatin operates in a less glamorous but arguably more consequential category — the software systems that help armed forces plan, supply, and sustain operations.
Why logistics software matters: Military effectiveness depends on supply chains as much as weapons. Ammunition, fuel, spare parts, medical supplies, and personnel movement all require planning systems that can adapt to changing conditions. Legacy military logistics systems are often decades old and poorly suited to contested environments where communications are degraded, and conditions change rapidly.
The commercial opportunity: Defense software is a market with characteristics that appeal to venture investors — long procurement cycles but large, sticky contracts; government buyers that are not subject to consumer market volatility; and demand driven by policy rather than consumer preference.
The deployment challenge: Software built for contested environments must function under conditions that most commercial software is not designed for — limited bandwidth, intermittent connectivity, and adversarial interference. That is a genuine technical constraint that raises the barrier to entry.
Editorial question: Does military logistics software become a durable venture category, or does it remain a specialized market dominated by established defense contractors with existing procurement relationships? Connect to: Defense Technology Startups · Tech M&A Watch · Enterprise AI Intelligence
Cybersecurity Startups Raise $5.26B in Q3 as AI Reshapes Security
The pattern: Global cybersecurity startup funding reached $5.26 billion in Q3 2026, a reported 78% year-over-year increase, with $12.7 billion raised through September, per The Wall Street Journal.
Where the capital is concentrated: Investment is flowing into AI security, identity, fraud prevention, and governance/risk/compliance. Each of these categories is being reshaped by AI in a specific way. AI security addresses the new attack surfaces that AI systems introduce — model inputs, inference infrastructure, agent permissions. Identity is being redefined as nonhuman identities (agents, service accounts, automated processes) multiply. Fraud prevention is increasingly adversarial, with AI systems on both sides. And GRC is expanding to cover AI governance, which regulators are beginning to formalise.
What the funding pattern means for founders: The concentration in these categories creates a split. Early-stage founders in AI security, identity and governance can raise at attractive valuations because they address genuine new problems. But founders in mature security categories — endpoint protection, network security, traditional SIEM — face a much harder environment, because those categories are dominated by incumbents and are not being fundamentally reshaped by AI.
The sustainability question: The 78% year-over-year increase reflects genuine enterprise demand — AI is creating new security problems that enterprises must solve. But it also reflects investor enthusiasm that may be outpacing actual market size. Not every AI security startup will become a durable business, even in a growing market.
Editorial question: Does the surge in cybersecurity funding reflect durable enterprise budgets for AI security, or is it a crowded category where many companies will be unable to build defensible businesses? Connect to: Cybersecurity Watch · AI Watch · VC Intelligence
AI Captures the Majority of Venture Capital as Mega-Rounds Multiply
The data: Venture-backed companies raised approximately $159 billion in Q3 2026, bringing the first three quarters to roughly $679 billion. The quarter posted a record number of billion-dollar funding rounds, per Crunchbase.
Where the capital is going: AI infrastructure, data centers, semiconductors, robotics, and energy are absorbing the overwhelming share of funding. These are not independent categories — they are interconnected layers of the same AI build-out. Data centers need power. AI systems need chips. Robots need edge compute. Energy infrastructure supports all of it.
What the concentration means: Headline funding totals can obscure a highly concentrated market. A $159 billion quarter does not mean that startups broadly are raising easily — it means that a relatively small number of capital-intensive AI and infrastructure companies are raising enormous amounts. The implications for non-AI startups: Companies outside AI, infrastructure, defense, and security face a materially harder fundraising environment. Investors have limited attention and capital allocation for categories that are not part of the dominant narrative. That is not a permanent condition — venture markets cycle — but it is the current reality.
The valuation question: Record round counts at high valuations create an inevitable question: do these companies generate returns that justify the prices? That question can only be answered over time, and the answer will determine whether the current funding pace continues or corrects.
Editorial question: Is the concentration of capital in AI and infrastructure a durable structural shift, or is it a cycle that will revert as returns come in and capital reallocates? Editorial note: The important questions are which companies receive capital, what milestones they must achieve, and whether returns justify the valuations — not the aggregate totals. Connect to: VC Intelligence · The Term Sheet · IPO Watch
Where Venture Capital Is Actually Moving
Today's rounds describe a market that is concentrating rather than broadening. Capital is flowing into categories where technical depth, capital intensity, and defensibility are highest — and away from categories where those characteristics are weaker.
| Category | Why Capital Flows There | Today's Example |
|---|---|---|
| AI agents | Autonomous systems that execute work, not just generate responses | Manus ($500M+) |
| Quantum computing | Deep technical moat; potential to solve problems classical computers cannot | Oratomic ($475M) |
| Open-source AI | Enterprise demand for customizable, controllable AI systems | Nous Research ($90M) |
| AI-native security | New attack surfaces created by AI adoption; non-discretionary enterprise spend | $5.26B in Q3 funding |
| Defense software | Policy-driven demand; large sticky contracts; less consumer-market exposure | Gallatin AI ($50M) |
| AI infrastructure | Underlies all of the above; capital-intensive and defensible | Q3 aggregate ($159B total) |
The direction is clear: capital is concentrating in categories with deep technical moats and non-discretionary demand. AI agents, quantum computing, open-source AI, AI-native security and defense software all share those characteristics. Categories without them — consumer apps, general SaaS, non-AI enterprise software — face a materially harder fundraising environment.
That concentration creates a strategic question for the market as a whole: if capital allocates into a narrow set of categories over multiple quarters, does the ecosystem lose the diversity that historically produced breakthrough companies in adjacent and unexpected areas? That is a question venture markets have historically answered by cycling, but cycles assume capital eventually reallocates, which requires returns to come in from current investments.
Funding Intelligence: What Investors Are Actually Buying
Three patterns define this week's funding activity:
1. Strategic investors are shaping which categories get funded. Nvidia, Microsoft's M12, and Samsung are participating in Nous Research's round. Bezos Expeditions is joining Oratomic's quantum round. Tencent and HSG are in Manus's round. These are not passive financial positions — they are strategic bets by established technology and industrial companies on which parts of the AI economy will matter. When Nvidia backs an open-source AI company and a quantum computing company in the same month, the composition of the cap tables tells you what those companies believe about the next decade.
2. Capital is concentrating in categories with non-discretionary demand. AI security, defense software, and AI infrastructure all have buyers who must spend regardless of economic conditions. AI security is non-discretionary because AI systems create new vulnerabilities that enterprises cannot ignore. Defense software is non-discretionary because governments have committed budgets. AI infrastructure is non-discretionary because every AI company needs compute. That combination — deep technical moat plus non-discretionary demand — is what attracts capital at scale.
3. Independence is becoming a strategic value in itself. Manus's $500M+ raise after unwinding its Meta acquisition is a useful signal: the market is willing to fund companies that choose — or are forced into — independence, provided the underlying technology is strong. That has implications for other companies considering whether to sell or remain independent. In a market where the largest acquirers are becoming more selective, the ability to raise independently at scale is a genuine strategic asset.
What to Watch Next
- Manus's valuation and IPO path. Watch whether the company confirms a valuation figure, and whether it begins the process of preparing for a public listing in the next 12–18 months.
- Oratomic's technical milestones. Track progress toward fault-tolerant quantum computing — the engineering milestones that will determine whether the valuation is justified by commercial readiness.
- Nous Research enterprise adoption. Monitor whether open-source AI agents see genuine enterprise deployment, or remain a technically interesting but commercially niche alternative.
- Defense software funding. Watch whether Gallatin's round catalyses more investment in military logistics and sustainment software, or whether the category remains specialized.
- Q4 cybersecurity funding. Track whether the 78% year-over-year increase continues, or whether investor enthusiasm moderates as the category becomes more crowded.
- Q4 venture capital totals. Watch whether the concentration of capital in AI and infrastructure continues at the same pace, or whether reallocation begins as returns come in from current investments.
- Non-AI startup fundraising. Monitor whether companies outside the dominant categories begin raising on more favourable terms, or whether the funding environment remains difficult for them.
Venture capital is concentrating in AI, infrastructure, security and defense — categories with deep technical moats and non-discretionary demand. The headline totals are extraordinary, but the distribution beneath them is narrow and getting narrower.
For founders: The funding question is no longer "Is this a good business?" It is: Does this business sit in a category where capital is currently allocating? AI agents, quantum computing, open-source AI, AI security and defense software are being funded. Consumer apps and non-AI enterprise software are not, at least not at scale. That is not a permanent condition, but it is the current reality.
For investors: Strategic investors — Nvidia, Microsoft, Samsung, Bezos Expeditions, Tencent — are increasingly shaping which categories get funded and at what valuations. Financial investors are competing with strategic capital for access, which raises the bar for participation in the most attractive rounds.
For enterprises: The vendors funded today define your options in the next procurement cycle. AI security startups funded now will provide the tools that govern agent behaviour. Quantum computing investments will eventually produce systems that solve problems currently out of reach. Open-source AI companies will offer alternatives to closed platforms. Track these companies as future vendors, not just as investment stories.
For the market: Record funding totals and record billion-dollar round counts are not the same as a healthy market. Concentration is a rational response to where returns currently appear highest — but it creates risk if the categories receiving capital do not produce the returns that justify current valuations. The next several quarters will begin to answer that question.
Recurring Format: Capital | Company | Round | Investors | Valuation | Use of Proceeds | Strategic Signal
Startup Funding Watch uses a consistent structured format so that capital flows can be tracked over time as a dataset rather than read as isolated news.
| Company | Round | Investors | Valuation | Strategic Signal |
|---|---|---|---|---|
| Manus | $500M+ | Boyu Capital, IDG Capital; Tencent, HSG, ZhenFund | Undisclosed | Independence after a failed acquisition is fundable |
| Oratomic | $475M Series B | ARCH, Spark, Khosla, Index, General Catalyst, Bezos Expeditions | ~$5.4B | Deep-tech moats attract capital at scale |
| Nous Research | $90M Series B | Robot Ventures (lead); Nvidia, M12, Samsung, 1789 Capital | ~$1.5B | Open-source AI attracts strategic investors |
| Gallatin AI | $50M Series A | 8VC, Silent Ventures (leads) | Undisclosed | Defense software beyond drones becomes fundable |
| Q3 Cybersecurity | $5.26B (quarter) | Various | N/A | AI security demand is non-discretionary |
| Q3 Global VC | $159B (quarter) | Various | N/A | Record concentration in AI and infrastructure |
The CODEW Stat
$679 billion — the total venture funding raised by venture-backed companies in the first three quarters of 2026, including roughly $159 billion in Q3 alone. It is an extraordinary number, and it is also a misleading one if read as a measure of market breadth. That capital is heavily concentrated in a narrow set of categories — AI, infrastructure, semiconductors, robotics, energy, security and defense. The headline total tells you how much is being deployed. The distribution beneath it tells you where the next generation of important companies is actually being built.
Sources: Data sourced from Reuters, TechCrunch, The Wall Street Journal, Axios, Crunchbase News, and The CODEW Funding Pulse, covering funding rounds, valuations, and investor activity reported on or before October 8, 2026. Reported but unconfirmed rounds and valuations are labeled accordingly. All factual claims regarding funding amounts, valuations, and deal terms are drawn from contemporaneous reporting and company disclosures. Editorial analysis is clearly distinguished from reported facts throughout. Round details and valuation figures should be verified against company announcements or additional independent reporting before publication. Rounds reported but not officially closed are labeled as reported or expected.
Reviewed by Erwin Castro
on
Friday, October 09, 2026
Rating:
No comments: