Venture Market 24 July 2026: Record $510 Billion for the Half-Year, Capital Concentration in AI, IPO and M&A.

Startup and Venture Capital News for 24 July 2026: Record $510 Billion in Half-Year, Capital Concentration in AI, Major Rounds, IPOs and M&A—An Overview for Investors

The global venture market enters the end of July 2026 in a state that is difficult to describe in a single word. Formally, this is the best half-year in the history of the industry: global venture investments reached a record $510 billion in the first half of 2026, surpassing the total for the entire year of 2025 ($440 billion). However, behind the record figure lies an unprecedented concentration: two companies—OpenAI and Anthropic—attracted $217 billion, or 43% of all venture dollars globally. For venture investors and funds, this indicates not a “boom,” but a restructuring of the logic behind capital allocation.

The key thesis of recent trading sessions and funding rounds is: investors are no longer paying for “exposure to AI” per se. They are paying for control over chokepoints—over infrastructure, regulated workflows, manufacturing capabilities, and systems that cannot be replaced with a mere API call. The deals announced this week demonstrate this logic with rare clarity.

Record Half-Year and the Price of Capital Concentration

The statistics for the first half of 2026 have rewritten all historical benchmarks for the venture market:

• $510 billion—global venture investments for H1 2026, compared to $375 billion in the peak half-year of 2021.
• $305 billion—the first quarter, the largest quarter in the history of the industry.
• $205 billion—the second quarter, distributed among more than 5,000 startups.
• Over 70% of capital in Q2 went to companies focused on artificial intelligence—compared to less than 50% a year earlier.
• 53% of Q2 volume consisted of meg rounds of $1 billion or more: 16 companies raised $108.6 billion.

For venture funds, this creates an uncomfortable arithmetic. A manager without allocation in OpenAI or Anthropic objectively showed weak performance in the half-year—not because they erred in selecting portfolio companies, but because the market benchmark was set by two capital tables. Late-stage funding increased by 141% year-on-year in Q2: capital did not broaden its reach, it deepened positions in already proven winners.

Exits Have Returned: A Record Quarter for IPOs and M&A

The most important news for LPs is not the volume of investment, but the recovery of liquidity. It is the exits, rather than paper re-evaluations, that pay the returns for vintage funds.

• 32 companies went public with valuations exceeding $1 billion in Q2 2026.
• SpaceX's IPO on June 12 was the largest venture company placement in history, raising $75 billion at a valuation of $1.77 trillion, with shares closing up 19%.
• 24 companies were acquired for prices starting at $1 billion, with a total volume of $113 billion, a record for all time.
• The acquisition of Anysphere (Cursor) by SpaceX for $60 billion is the largest startup acquisition in market history.
• The next largest placements after SpaceX were the inference chip manufacturer Cerebras Systems and the quantum company Quantinuum.

Years of backlog in exit queues have finally begun to dissipate. For late-stage investors, this fundamentally alters the risk calculation: private capital is again converting into real liquidity, not just headline valuations.

Cybersecurity: The Category with the Highest Conviction Conversion

Cybersecurity remains a domain where venture funds are willing to back growth until revenue is disclosed. The company Glow emerged from stealth mode with a $180 million Series A round at a valuation of $1.2 billion. The syndicate included Sequoia Capital, Cyberstarts, Greenoaks, Redpoint Ventures, Index Ventures, Lux Capital, and Operator Collective.

Glow's thesis is simple and therefore compelling: the end device becomes the main attack surface in an era where employees launch AI agents, install developer tools in minutes, and introduce risk faster than security can respond. The company does not aspire to be just another detection layer alongside CrowdStrike, Microsoft, SentinelOne, and Palo Alto Networks, but rather aims for a political and orchestration level, determining which software and agents are allowed to enter the perimeter.

In the same segment, StrongestLayer raised $4.1 million led by Inovia Capital, bringing its seed funding to $9.3 million. The company builds mail protection based on intuition about message intent rather than signatures and reputation databases—an answer to the rise of BEC attacks that contain no overtly malicious payloads.

Defence Technologies: Geopolitics as an Investment Thesis

The most politically charged deal of the week was the round from Cathedral: $160 million at a valuation of $1.4 billion co-led by Andreessen Horowitz and Sequoia Capital. The startup, founded by alumni of the Department of Government Efficiency, is developing AI systems for military cyber operations—both defensive and offensive—and is reportedly exploring acquisition or partnership possibilities for dedicated computing resources.

For venture investors, Cathedral illustrates three converging forces: national security software based on AI, direct connections of founders with federal procurement circles, and capital conviction that geopolitical competition justifies aggressive underwriting. The flip side is political risk: proximity to power accelerates contracts, but makes the company vulnerable to changing political conditions.

Physical AI and Robotics: From Demonstrations to Unit Economics

Robotics attracted $18.8 billion since the beginning of 2026—more than the total for all of 2025. The key change lies in the nature of founders' arguments: buyers are interested not in demonstrations but in throughput, uptime, and cost per unit.

• Humanoid (London) — $152 million Series A at a post-money valuation of $1.35 billion led by Prime Movers Lab with participation from Schaeffler, Bosch, Fubon Financial Holding Venture Capital, and Aglaé Ventures. The total amount of raised capital is $270 million. Partnerships with SAP, NVIDIA, Bosch, and Siemens alongside a commercial agreement with Schaeffler are transitioning the project from prototypes to large-scale deployment. The company positions the round as proof that Europe can grow a globally competitive player in physical AI.
• Gritt — $26 million Series A led by Obvious Ventures with participation from Union Square Ventures and Active Impact Investment. The company automates solar installation: a team of eight installs about 800 panels a day in the traditional way compared to 3,000–4,000 using Gritt's systems. The contracted volume is 2.8 GW over the next 18 months.
• 1872 (Cincinnati) — $15 million seed round from The O.H.I.O. Fund. The founders are former SpaceX engineers building an autonomous steel structure factory in partnership with Path Robotics.

Energy and Materials: Supply Chain Sovereignty as an Asset Class

The company Sila raised $300 million led by Atreides Management and Sutter Hill Ventures with participation from 8VC, Bessemer Venture Partners, Matrix Partners, and funds managed by T. Rowe Price. The total amount of funding reached approximately $1.6 billion. The funds will be used to expand the production of silicon-carbon anodes in Moses Lake, Washington.

The investment thesis here extends beyond the electric vehicle market: Sila sells its technology to drones, satellites, electronics, robotics, and AI systems simultaneously. Capital is seeking “picks and shovels” capable of riding multiple demand curves—especially where data centres and defence procurement drive up demand for batteries.

Of particular note is Bluecore Energy—a pre-seed round of approximately $10 million led by Slauson & Co. The company is developing small modular reactors with water cooling on floating barges and has already delivered the first barge with a test reactor to Long Beach port. The initial system, rated at 10 MW, is designed to power the equivalent of 15,000 households or a large port. AI’s appetite for electricity has become a standalone engine for startup formation.

Fintech: Fewer Deals, More Infrastructure

Global fintech funding grew about 23% year on year in the first half of 2026, despite a more than 25% drop in the number of deals. Capital is concentrating on large infrastructure bets.

1. Augustus — $180 million Series B at a valuation of $1 billion led by Tiger Global with participation from Hummingbird and QED. The company is building a “Global Dollar Bank”—providing direct access for international fintechs and banks to dollar accounts and clearing rails through a federally-chartered institution with conditional approval from OCC. In total, $210 million has been raised.
2. Cashea (Caracas) — $100 million disclosed in a single announcement: Series A of $40 million led by Spice Expeditions (March 2026) and Series B of $60 million led by FinSight Ventures (June 2026). More than 10 million consumer accounts, 40,000 stores, and over 100 million transactions. The case demonstrates: frontier geographies receive funding if the company shows local density and borrower payment discipline.

Healthcare and Biotech: Capital Has Become Disciplined

Biotech funding has divided into two distinctly different segments. Late-stage, clinically de-risked assets still gather oversubscribed rounds; early projects are funded only under a narrow, specific technical wedge.

• Crystalys Therapeutics — $130 million Series B led by Frazier Life Sciences with participation from Wellington Management, HBM Healthcare Investments, Soleus Capital, Cormorant Asset Management, Novo Holdings, and SR One. In total, $335 million. The funds will support the third phase and preparation for the commercialization of the drug dotinurad against gout.
• Candid Health — $120 million Series D led by Sixth Street Growth with participation from Oak HC/FT, 8VC, and Y Combinator. The company automates the cycle of medical billing—a segment that burns about $280 billion annually within the US healthcare system. The valuation has tripled relative to Series C, with a 190% year-on-year growth in contracted annual revenue and a net dollar retention of 180%.
• Tikva Allocell (Singapore) — $8 million Series A from Kantharos Capital for an IND application filing by year-end.
• Brenus Pharma (Lyon) — €11 million Series A extension, totaling €38 million, with participation from Bpifrance, Sambrinvest, and Korea Omega Investment Corp.
• Immitra Bio (Zurich) — €2.58 million pre-seed led by Backbone Ventures and OCCIDENT to advance in-vivo genome editing.

Second-Order AI Infrastructure: Orchestration Instead of Models

A distinct emerging class of deals involves companies making already built AI infrastructure suitable for industrial use. Meshy raised nearly $400 million in Series B at a valuation of $1.5 billion—the largest disclosed round in the AI-3D segment; the company's products are used by teams within five of the ten largest tech corporations globally, with ARR increasing approximately twelve-fold year on year. SkyPilot emerged from stealth mode with $20 million in seed funding led by Lux Capital with participation from Amplify Partners, Coatue, and Foundation Capital: the company integrates fragmented computing resources—hyperscalers, neo-clouds, Kubernetes clusters, and various types of accelerators—into a single management layer.

The British company CuspAI earlier this week closed a Series B round of $450 million with support from Kleiner Perkins, NEA, Bezos Expeditions, the UK government, AMD Ventures, and Lux Capital, bringing the total raised to over $650 million. The focus is on AI for discovering new materials.

What This Means for Venture Funds and Institutional Investors

Practical takeaways for asset managers as of the end of July 2026:

• Record volumes do not equal a broad market. With $510 billion for the half-year, 43% went to two companies. When evaluating portfolio returns, it is more accurate to use median rather than weighted benchmarks.
• The quality of the syndicate has become a signal of survival. The market rewards the presence of specialised lead investors capable of supporting a company in subsequent rounds, affecting the price as much as metrics do.
• Security is determined by control rather than technology. Production assets, regulatory licenses, built-in distribution, and workflow data are what survive the commoditisation of models.
• The exit window is open, but selectively. Record IPOs and M&A in Q2 provide late investors with grounds for exit, but the public market accepts companies that look like infrastructure rather than functions.
• Geography has given way to categories. The US share dropped from 83% in Q1 to two-thirds in Q2—a preliminary signal of capital redistribution towards Europe and Asia.
• Capital efficiency has returned to the agenda. Companies demonstrating growth with small teams and positive unit economics are receiving valuation premiums that did not exist in the "growth at any cost" cycle.

Conclusion: The Market is Narrow, but Open

The venture market at the end of July 2026 is neither overheated nor closed. It is narrow, strategic, and increasingly intolerant of abstractions. Large cheques are still being written—but more frequently reserved for companies that do not appear as experiments but as future infrastructure for specific segments of the economy. For venture investors and funds, the primary skill of this new cycle lies in distinguishing between a company controlling a chokepoint and a company selling a function layered over someone else's model. This distinction, rather than the growth rate of the AI industry, will define the returns of the 2026 vintage.