Cost Reduction and Efficiency Boosting AI: VC Spending More and More

Author: Zen, PANews

On August 11, the veteran venture capital firm Accel announced the completion of a new fundraising round of $3.5 billion. In April this year, the company had just raised a $5 billion late-stage investment fund, and in just four months, Accel has loaded $8.5 billion into the chamber.

Accel views AI as a "super cycle" of technology that is still in its early stages. In its judgment, AI is significantly compressing the time cycle for startups from product conception to scaling expansion.

However, contrasting with the narrative of "cost reduction and efficiency enhancement," the AI primary market is becoming increasingly expensive. Seed financing scales continue to rise, early valuations are constantly increasing, and some AI companies that have not yet formed mature products and revenue models are already able to obtain capital that was previously only available to growth-stage companies.

AI has reduced some costs of starting a company but has simultaneously pushed up the prices for acquiring equity in quality AI companies.

Decreasing Startup Costs, Increasing Capital Stakes

AI tools are enabling some software, SaaS, and fintech startups to complete product development and early validation with less capital. A limited-sized team can now accomplish tasks that previously required more engineers, sales, and operational personnel.

According to research data from equity management service provider Carta, the median team size of seed-stage startups in the U.S. is currently only 4 people; the average number of employees in Series B has decreased from 53 in 2023 to 45, and Series D has dropped by 29% from its peak to 131 people.

In terms of the financing structure of startups, AI is driving the primary market towards polarization, with the financing market gradually showing a clear "barbell" structure.

On one hand, for lightweight startups that rely on AI tools to significantly reduce fixed costs, the initial capital required for product development and business validation is clearly decreasing; on the other hand, a few startups with top teams and technical backgrounds are beginning to see a comprehensive increase in financing scale, obtaining financing and valuations far higher than ordinary startups in their early stages.

Carta's statistics show that in the first quarter of 2026, about 3,000 startups in the U.S. completed Pre-Seed financing, with the final financing scale expected to be about $2.9 billion, roughly on par with the previous few quarters. Among them, AI startups accounted for 50% of the funds raised, compared to about 30% a few years ago.

In terms of fund distribution, the proportion of medium-sized financing between $1 million and $2.5 million has dropped from 24% in the first quarter of 2023 to 18%, while smaller financing below $1 million has become more common, and the proportion of large financing above $2.5 million has remained stable.

Moreover, the valuations of leading projects are further widening the gap. In SAFE transactions with financing scales exceeding $2.5 million, the top 10% of startups by valuation have already reached a valuation ceiling of over $100 million; in the second quarter of 2026, the top 5% of seed round financing projects have reached about $200 million, a 177% increase from $72.2 million in the same period of 2025.

AI is redefining the traditional concept of "early-stage financing," and this change has been particularly evident in recent months.

A group of core researchers and executives from leading AI companies like Google and OpenAI have just left to start their own ventures, and although these projects are still in the very early stages, capital pricing has rapidly entered the hundreds of millions or even billions of dollars.

In early August, Jeff Dean, who worked at Google for nearly 27 years, along with several core researchers from Google and DeepMind, including Sanjay Ghemawat, Quoc Le, and Oriol Vinyals, left to establish an AI scientific research company called Discovery Loop.

When the company was officially established, it had not even completed team and office setup, but it had already received support from institutions such as Radical Ventures, Khosla Ventures, Lightspeed, and Kleiner Perkins, with Alphabet also participating as a founding investor. Subsequently, reports indicated that Discovery Loop was discussing a financing round of about $1 billion, corresponding to a valuation of about $10 billion.

Kevin Weil, the former Chief Product Officer of OpenAI, is preparing an AI scientific company whose name and product have not yet been disclosed since leaving this year. According to the latest reports from August, this project is seeking about $150 million in financing, corresponding to a valuation of at least $750 million.

In April of this year, Ineffable Intelligence, founded by former DeepMind reinforcement learning core researcher David Silver, completed an $1.1 billion seed round, achieving a valuation of $5.1 billion, becoming one of the largest seed financings in Europe.

In the entrepreneurial version of the AI era, capital is turning the future potential of top teams into "one-time realizations."

Rising Valuations Push Up VC's Holding Costs

In the past, product capability, user growth, and revenue were often important bases for gradually increasing valuations. Today, for teams from top labs like OpenAI and Google DeepMind, research backgrounds, talent combinations, and the potential to build platform companies in the future can already be capitalized at the early stages of the company.

Venture capital has a very simple business model, where returns largely depend on entry prices and holding ratios. For early-stage funds, establishing a sufficiently large initial position when company valuations are still low and maintaining the corresponding equity ratio in subsequent financing processes is crucial for obtaining excess returns.

However, in the current AI investment boom, the window for acquiring quality project equity at low prices is rapidly closing.

The latest data released by Carta in July further shows that in the past six months, the equity dilution ratio in financing for U.S. software companies has continued to decline. The median dilution ratios for seed and Series A rounds are both around 18%, while Series B has dropped to 12%, and Series C is below 10%. During the same period, the median valuation for seed rounds reached $24.3 million, while Series A and B reached $80 million and $191 million, respectively.

In other words, while financing scales and company valuations continue to rise, the equity ratio given up by founding teams has not increased in tandem. For VCs hoping to obtain or maintain a high holding ratio, the capital required to acquire the same proportion of equity is significantly increasing.

Assuming a startup has a pre-financing valuation of $90 million, a VC investing $10 million could roughly acquire 10% of the shares. If the valuation of the same stage company rises to $490 million, then the investment amount required to obtain close to 10% equity will increase to about $50 million.

This also means that VCs need to not only invest but also keep up with the pace. This requires large VCs to possess two capabilities: to secure sufficient initial holdings in the early stages and to reserve ample funds for subsequent financing after the rapid rise in company valuations.

This trend has become a consensus among all leading VCs. In the $3.5 billion fundraising mentioned at the beginning, Accel has planned a $1.35 billion global expansion fund, which is primarily aimed at supporting larger initial investments in early stages and rapid follow-ups. The $5 billion late-stage fund raised in April provides capital for Accel to continue its investment capabilities as portfolio companies enter the growth stage.

In January of this year, a16z completed a fundraising of over $15 billion in one go, with $6.75 billion allocated for growth-stage investments and $1.7 billion specifically targeting AI infrastructure.

B Capital subsequently completed a $500 million early-stage fund, doubling the scale of its previous similar fund. Its management pointed out that as a large amount of capital enters the early market, some early financing valuations and transaction scales are increasingly approaching those of past growth stages.

Matthew Effect: Capital Concentrating Further on a Few Leading AI Projects

Funds in the primary market are further concentrating on a few leading projects.

The latest data from Carta shows that in the first half of 2026, the companies it covers completed $58.7 billion in venture financing, up from $56.5 billion in the same period last year, but the differentiation between different financing stages has significantly widened. Among them, seed round financing dropped from $6.5 billion to $3.8 billion, Series B dropped from $13.5 billion to $10.4 billion, while Series A remained roughly stable at $12.7 billion; financing in Series C and later stages grew from $23.9 billion to $31.8 billion.

Crunchbase statistics show that in the second quarter of 2026, over 70% of global startup financing flowed to AI companies, a significant increase from less than 50% in the same period last year. During the same period, 16 companies completed financing of over $1 billion, totaling $108.6 billion, accounting for 53% of the total venture capital for the quarter.

The capital-absorbing effect of leading companies is even more pronounced. In the first half of this year, OpenAI and Anthropic together received $217 billion in financing, accounting for 43% of the total financing for global startups. This means that the current AI investment boom is not about capital being evenly distributed among a large number of startups, but rather increasingly concentrated on a few foundational model companies and already market-recognized leading projects.

This concentration further intensifies the competition among large funds for leading projects.

If only a few AI companies can ultimately form global platforms, then large funds missing out on these companies' shareholder lists may directly impact the return performance of the entire fund cycle. Compared to expanding the number of investment projects, more and more institutions are choosing to reduce the number of investment targets and invest larger amounts in a few high-certainty projects.

This also creates a self-reinforcing mechanism—quality AI companies grow faster, leading VCs to enter the competition earlier. The resulting competition raises early valuations, and the capital required to maintain equity ratios increases accordingly. Subsequently, capital further concentrates on leading projects, reinforcing the advantages of large funds in subsequent financing.

Ultimately, AI not only expands the financing gap between startups but is also redefining the competitive landscape of the VC industry itself. Smaller funds are particularly affected.

A $100 million fund that could previously diversify investments across dozens of seed projects may find its ability to participate in popular AI projects significantly diminished if a leading AI project reaches tens of millions of dollars in a single round of financing while also requiring investment institutions to reserve funds for subsequent financing.

Large institutions can cover a company's entire capital cycle through funds at different stages: early-stage funds are responsible for establishing holdings, growth funds continue to add, and late-stage funds further maintain equity ratios. Accel's raising of $8.5 billion in four months is a typical reflection of this trend.

High Valuations Are Pricing in Future Growth Expectations

However, the expansion of VC fundraising does not mean that achieving investment returns will be easier. In fact, the higher the entry valuations, the higher the requirements for future growth and exit scales of companies.

If a startup receives investment at a $100 million valuation, growing to $1 billion can bring a tenfold valuation increase; but if a company's early valuation has already reached $10 billion, achieving the same tenfold growth would require a final valuation of $100 billion.

Thinking Machines' first round financing valuation has already reached $12 billion; SSI's valuation has risen to $32 billion in less than a year since its establishment. These valuations reflect the scarcity of top AI teams, but also mean that a considerable portion of future growth expectations has already been priced in.

The current risks in the AI primary market lie here. If a few platform companies emerge in the future with revenue and profit scales sufficient to support valuations in the hundreds of billions, then today's high valuations may still be digestible; but if most AI companies ultimately fail to establish sufficiently strong technical and business barriers, the excessively high entry prices will directly compress the potential return space for investment institutions.

Currently, this investment model still performs well on paper. Carta's data shows that some VC funds established in 2023 and 2024 currently have internal rates of return that outperform some older funds from 2017 to 2020. However, a considerable portion of these returns comes from valuation re-evaluations brought about by subsequent financing, rather than actual cash returns generated from IPOs or mergers and acquisitions. The continuously rising valuations in the primary market today still need to be validated through future revenue growth and exit prices.

But for large VCs like Accel, the more realistic issue at hand is not waiting for valuations to return, but how to avoid missing out on the few winners that may emerge in this technological cycle.

This is also one of the reasons for the continuous expansion of large funds. The growth and financing cycles of AI companies are clearly shortening, and once a project gains market recognition, its valuation may rapidly increase over several rounds of financing. If VCs wish to secure sufficient equity in the early stages and maintain their equity ratios in subsequent financing, they need to prepare capital reserves far exceeding those of the past in advance.

Accel defines AI as a "super cycle" of technology that is still in its early stages. Under this judgment, the $3.5 billion early and expansion fund and the previously raised $5 billion late-stage capital actually correspond to the same strategy: to enter potential winners as early as possible and reserve space for continuous follow-up investments after their valuations rise rapidly.

This also constitutes the most obvious contradiction in the current AI investment boom: AI has reduced some costs of entrepreneurship but has not lowered the costs of investing in quality AI companies.

For VCs, what has truly become expensive is not just the funds needed to support a startup's growth, but the prices required to acquire and maintain sufficient equity ratios in quality projects within an increasingly shorter window.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.