Key Developments on 29 July 2026: The Numbers Shaping the Agenda
Below are the benchmark figures around which the current market discussion is built:
- $510 billion — global venture investments in the first half of 2026; Q1 delivered $305 billion, Q2 added a further $205 billion across more than 5,000 companies.
- 43% — the combined share of two companies, OpenAI and Anthropic, in global venture funding for the half-year ($217 billion in total).
- Over 70% — the share of AI startups in global venture investments during the second quarter, compared with approximately 50% a year earlier.
- $412.7 billion — venture investments in the US over the half-year, of which $355.9 billion (86%) went to AI companies.
- $251 billion — raised from 86 US IPOs since the start of the year, more than five times the total for the whole of 2025 ($47.4 billion).
- $113 billion — the volume of startup acquisitions valued at $1 billion or more in the second quarter, a record for the entire history of tracking.
- 5.09 billion rubles — the volume of the Russian venture market for the half-year, down 40% year-on-year with the number of deals halved.
A Half-Year Record: Why $510 Billion Does Not Mean "The Market Is Back"
The record volume of venture funding was not driven by a widening of the funnel, but by a few mega-rounds. The number of deals in the first half remained virtually flat, while in Asian markets the transaction count fell to multi-year lows despite record amounts. In other words, the average cheque size has multiplied, while access to capital has narrowed.
Late-stage funding in the second quarter rose by approximately 141% year-on-year. This is a fundamental shift in venture fund behaviour: capital is flowing not into expanding portfolios with new names, but into recapitalising already proven leaders. For fund managers, this means a more predictable but less asymmetric return profile; for LPs, it implies increasing correlation between funds pursuing different strategies.
Capital Concentration: The Key Risk on the Agenda
A situation in which two companies absorb 43% of global venture capital in a half-year has no historical precedent. Added to this is a geographic imbalance: around 88% of all AI startup investments go to companies headquartered in the US. At the same time, the US share of total Q2 volume declined from 83% to 66–67% — capital is simultaneously concentrating by sector and internationalising by geography.
This raises three practical questions for investment committees:
- How diversified is a fund’s portfolio if the bulk of sector returns are determined by just a few private companies?
- How should second-tier AI startups be valued when valuation benchmarks are set by rounds of unprecedented scale?
- What will happen to the entire sector’s multiples if even one of the leaders disappoints the public market?
Late-July Deals: Where the Money Actually Went
The final week of July provided a telling snapshot of venture fund priorities. The most notable funding rounds include:
- Etched — $300 million, Series C, inference chips, led by Sequoia.
- CuspAI — $450 million, Series B, AI for new materials development (Kleiner Perkins, NEA).
- Meshy — approximately $400 million, Series B, 3D content generation.
- Glow — $180 million, Series A, cybersecurity (Sequoia, Cyberstarts).
- Cathedral — $160 million, defence cyber-AI (Andreessen Horowitz, Sequoia).
- Humanoid — $152 million, Series A at a $1.35 billion valuation; Europe’s first humanoid robotics unicorn.
- Neo — $100 million upon exiting stealth mode, application security in the age of AI agents.
Earlier in July, the market witnessed even larger transactions: $1.8 billion for defence-focused Helsing, $1 billion for SambaNova Systems, $800 million for Together AI, $700 million for medtech platform Neko, and €411 million for fusion energy project Proxima Fusion. The overarching conclusion: venture capital is funding not so much applications as the "operating system" of the new economy — computing, energy, security, and industrial robotic systems.
Physical AI, Defence, and Deep Tech: The New Map of Priorities
Three themes are shaping investment fashion in the second half of 2026. The first is physical AI: models integrated with hardware, from construction robots to industrial perception systems. The second is defence and sovereign technologies, where European startups are competing with their US counterparts on cheque sizes for the first time in a decade. The third is energy for data centres: fusion, geothermal, and grid projects are being funded as an infrastructure asset class rather than a venture one.
It is also notable that cybersecurity has become a derivative of the proliferation of AI agents: investors are funding companies that solve problems created by generative models themselves. This is a sustainable "second-order" pattern and will remain a source of deals at least through year-end.
The 2026 IPO Window: Open, But Not for Everyone
The primary market is experiencing its strongest comeback since 2021. By the end of July, the US had seen 86 IPOs totalling $251 billion; global proceeds for the half-year reached $178 billion (+205% year-on-year) across 524 deals. Technology offerings delivered an average first-day pop of 44.5%, and the combined valuation of companies in the IPO pipeline exceeded $2.1 trillion.
Yet the structure of this record is as concentrated as the venture landscape. SpaceX’s $85.7 billion listing at a $1.75 trillion valuation accounted for roughly a third of all capital raised this year. Anthropic filed on 1 June after a $65 billion round; OpenAI filed confidentially on 8 June at a private valuation of $852 billion. Strava is preparing a listing at around $2.2 billion. Meanwhile, Databricks publicly declined a 2026 listing in favour of 2027, discussing a private round at a $165–175 billion valuation versus $134 billion six months earlier. Canva and Cohere are still viewed by the market as 2027 candidates.
M&A and Exits: The Best Quarter in Five Years
For the first time since 2021, exit momentum has caught up with funding momentum. In the second quarter, 32 companies went public with valuations above $1 billion, while another 24 were acquired for $1 billion or more, totalling a record $113 billion. For venture funds, this means an unlocking of DPI: LP distributions have finally begun to return to levels that allow a full cycle of new fund over-subscription.
Nonetheless, the quality of exits remains uneven. Large strategic acquisitions are concentrated in AI infrastructure, semiconductors, and biotech, whereas mid-sized, classical SaaS companies are still exiting at a discount to their 2021 round valuations.
Fundraising and Dry Powder: Capital Is Available, But Access Is Restricted
At the global level, private markets hold approximately $3.9 trillion in uncommitted capital, of which around $600 billion is directly attributable to venture funds. At the same time, the share of successfully closed funds has fallen to about 57%, compared with 94% in 2020 — LPs have become notably more selective and prefer proven platforms over new managers.
The practical market consequence is that the gap between top-quartile funds and the rest continues to widen, while emerging managers increasingly access deals through syndicates, SPVs, and co-investments with large platforms.
Russia and the CIS: A Market in Severe Selection Mode
The Russian venture market is moving in the opposite direction to the global trend. In the first half of 2026, venture investments totalled 5.09 billion rubles — 40% less than a year earlier. There were 50 deals, half the number of the first half of 2025, with an average cheque of 113.2 million rubles. The largest share of investments went to artificial intelligence and machine learning — the sectoral focus mirrors the global trend, but the scale does not.
Industry analysts compare current indicators to levels seen in 2009–2011. The logic of funding has changed structurally: with a high key interest rate, deposits and the debt market compete with venture returns, so investors demand confirmed revenue, positive unit economics, and a clear path to profitability from startups — not just a "promising idea." Corporate venture capital, sector-specific funds, and club syndicates remain the main sources of capital.
Conclusions for Venture Investors and Funds
The agenda for 29 July 2026 boils down to four key points:
- Record ≠ Broad market. The aggregate $510 billion masks a narrowing funnel: capital is available to category leaders, not to the average startup.
- Concentration is a standalone risk. Portfolios whose returns depend on a few AI leaders require stress-testing for the scenario of a disappointing debut by one of them.
- The exit window is open, but selective. Companies with valuations of $2–5 billion, sustainable revenue, and proximity to profitability have a real chance to use the current IPO cycle.
- Infrastructure bets trump application bets. Computing, energy, security, and physical AI offer a more defensible position than applications built on top of others’ models.
The market has entered a phase where an abundance of capital coexists with a scarcity of access to it. For venture funds and institutional investors, this means a return to fundamental discipline: selection quality, valuation rigour, and sober liquidity planning — regardless of how impressive the headline figures for the half-year may look.
