Author: Zen, PANews
On August 11, established venture capital firm Accel announced the completion of a new funding round of $3.5 billion. Just four months after raising a late-stage investment fund of $5 billion in April, Accel has loaded $8.5 billion into the chamber.
Accel views AI as a technology "super cycle" that is still in its early stages. In their judgment, AI is significantly compressing the time span between startup product conception and large-scale expansion.
However, contrasting with the narrative of "cost reduction and efficiency improvement," the primary market for AI is becoming increasingly expensive. Seed financing scales continue to rise, early valuations keep increasing, and some AI companies that have not yet developed mature products or revenue models are already able to access capital that was previously only available to growth-stage firms.
AI has reduced some costs associated with starting a company, but at the same time, it has driven up the price of acquiring equity in quality AI companies.
Startup costs are down, capital bets are up
AI tools are enabling some software, SaaS, and fintech startups to complete product development and early validation with less capital. A limited-sized team today can accomplish what previously required more engineers, sales, and operations personnel.
According to research data from equity management service provider Carta, the current median team size for seed-stage startups in the U.S. is only 4 people; the average employee count for Series B has decreased from 53 in 2023 to 45, while Series D has dropped 29% from its peak to 131.
In the financing structure of startups, AI is driving the primary market towards polarization, with the financing market increasingly exhibiting a clear "barbell" structure.
On one hand, for lightweight startups that significantly reduce fixed costs with the aid of AI tools, the initial capital required for product development and commercial validation is clearly decreasing; on the other hand, a small number of AI startups with top talent and technological backgrounds are beginning to see a comprehensive rise in financing scales, securing funding and valuations far exceeding those of typical startups in their early stages.
Carta statistics show that in the first quarter of 2026, around 3,000 startups in the U.S. completed Pre-Seed financing, with an expected total financing scale of about $2.9 billion, roughly on par with past quarters. AI startups accounted for 50% of the funds, compared to about 30% a few years ago.
In terms of funding distribution, medium-sized financing between $1 million and $2.5 million has dropped from 24% in the first quarter of 2023 to 18%, while smaller financings under $1 million have become more common; large financings over $2.5 million have remained stable.
Furthermore, the valuations of top projects are widening further apart; among SAFE transactions with financing over $2.5 million, the top 10% of startups have reached valuation caps exceeding $100 million; by the second quarter of 2026, the top 5% of seed financing valuations have reached about $200 million, a 177% increase from $72.2 million in the same period of 2025.
AI is redefining "early financing" in the traditional sense, a change that has been especially evident in recent months.
A group of core researchers and executives from leading AI companies like Google and OpenAI recently left to start their ventures, with their projects still in the very early stages, but capital pricing has rapidly entered the hundreds of millions and even billions of dollars.
In early August, chief scientist Jeff Dean, who worked at Google for nearly 27 years, together with several core researchers like Sanjay Ghemawat, Quoc Le, and Oriol Vinyals, left to establish an AI research company called Discovery Loop.
When the company was officially founded, it had not even completed building its team and office space, yet it had already received support from Radical Ventures, Khosla Ventures, Lightspeed, Kleiner Perkins, and Alphabet, which also participated as a founding investor. Reports later indicated that Discovery Loop is discussing about $1 billion in funding, corresponding to a valuation of around $10 billion.
Former chief product officer at OpenAI Kevin Weil, after leaving this year, is preparing to found an AI science company whose name and products are not yet public. According to the latest reports in August, this project is seeking about $150 million in funding, corresponding to a valuation of at least $750 million.
In April of this year, David Silver, a former core researcher in reinforcement learning at DeepMind, established Ineffable Intelligence, which completed an $1.1 billion seed round, reaching a valuation of $5.1 billion and becoming one of the largest seed financings in Europe.
In the entrepreneurial version of the AI era, capital is liquidating the future potential of top teams at once.
Valuations rise, driving up VC ownership costs
In the past, product capability, user growth, and revenue were often critical grounds for gradual valuation increases. Now, for teams from top labs like OpenAI and Google DeepMind, research credentials, talent composition, and the potential to build platform-type companies can already be capitalized at the company's early stages.
Venture capital has a very simple business model, where returns largely depend on entry prices and ownership proportions. For early-stage funds, establishing a sufficiently large initial stake when the company's valuation is still low and maintaining the corresponding ownership ratio during subsequent financing rounds is crucial for realizing excess returns.
However, in the current AI investment boom, the window for acquiring quality project equity at low prices is rapidly closing.
Latest data released by Carta in July shows that, over the past six months, the equity dilution rate in U.S. software company financing continues to decline. Among them, the median dilution rates are about 18% for seed and Series A, 12% for Series B, and below 10% for Series C. Meanwhile, the median seed valuation has reached $24.3 million, with Series A and B reaching $80 million and $191 million, respectively.
In other words, while the funding scale and company valuations continue to rise, the equity percentages given up by founding teams have not increased correspondingly. For VCs looking to obtain or maintain a high ownership stake, the capital required to acquire the same proportion of equity is significantly increasing.
Assuming a startup has a pre-funding valuation of $90 million, and the VC invests $10 million, they could roughly acquire 10% equity. If the same company's valuation rises to $490 million, then acquiring close to 10% equity would require an investment of about $50 million.
This also means that VCs need to not only invest but also keep up with rising valuations. This requires large VCs to possess two abilities: to secure enough initial holdings in the early stages and to reserve sufficient capital for subsequent financing after the company's valuation rises rapidly.
This trend has become a consensus among leading VCs. In the previously mentioned Accel's $3.5 billion fundraising, a $1.35 billion global expansion fund was planned, with a significant purpose being to support larger early-stage first investments and rapid follow-ons. The $5 billion late-stage fund raised in April provides capital for Accel to maintain investment capabilities after portfolio companies enter the growth phase.
In January of this year, a16z completed over $15 billion in fundraising at once, with $6.75 billion allocated for growth-stage investments and $1.7 billion specifically targeting AI infrastructure.
Subsequently, B Capital completed a $500 million early-stage fund, doubling the scale of its previous similar fund. Their management pointed out that after a large influx of capital into the early market, some early financing valuations and transaction sizes are becoming increasingly similar to those from past growth stages.
Matthew effect, capital further concentrates on a few top AI projects
Capital in the primary market is further concentrating on a few top projects.
Latest data from Carta indicate that in the first half of 2026, the companies they cover completed $58.7 billion in venture financing, up from $56.5 billion in the same period last year, but there is a significant widening of the divide between different financing stages. Specifically, seed funding has decreased from $6.5 billion to $3.8 billion, Series B from $13.5 billion to $10.4 billion, and Series A remained steady at $12.7 billion; while funding from Series C and later phases has grown from $23.9 billion to $31.8 billion.
Crunchbase statistics show that in the second quarter of 2026, over 70% of global startup funding flowed to AI companies, a clear increase from less than 50% in the same period last year. During this time, 16 companies completed financing of over $1 billion, totaling $108.6 billion, accounting for 53% of the entire quarter's venture capital amount.
The money-raising effect of top companies is even more apparent. In the first half of this year, OpenAI and Anthropic together raised $217 billion, making up 43% of total startup financing globally. This indicates that the current AI investment boom is not capital evenly spreading across many startups, but is increasingly concentrating on a few foundational model companies and top projects that have already gained market recognition.
This concentration further intensifies the competition among large funds for top projects.
If only a few AI companies can ultimately form global platforms, then large funds failing to get on the shareholder list of these companies may directly impact the fund's overall performance over the investment cycle. Rather than expanding the number of investment projects, an increasing number of institutions are choosing to reduce the number of investment targets and invest larger sums in a few high-certainty projects.
This has also created a self-reinforcing mechanism—high-quality AI companies are growing faster, prompting VCs to enter the competition earlier. Increased competition drives up early valuations, thereby increasing the capital needed to maintain ownership ratios. Subsequently, capital further concentrates on top projects, and large funds are reinforced in subsequent financing.
Ultimately, AI is not only widening the financing gap between startups but is also reshaping the competitive landscape within the VC industry itself. Smaller funds are particularly impacted by this trend.
A $100 million fund, in the past, could diversify investments across dozens of seed projects. However, if a top AI project requires tens of millions in a single financing round, and investment institutions also need to reserve funds for subsequent financing, then the capacity of small funds to participate in popular AI projects will significantly decrease.
Large institutions can cover a company's entire capital cycle through different stages of funds: early funds handle the establishment of holdings, growth funds continue to follow on, and late funds maintain ownership 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 enlarged fundraising scales of VCs do not necessarily mean that investment returns will be easier to achieve. In fact, the higher the valuation, the higher the demands are for the future growth and exit scales of the company.
If a startup receives investment at a valuation of $100 million, growing to $1 billion brings a tenfold valuation increase; but if a company's early valuation already reaches $10 billion, achieving the same tenfold growth would require a final valuation of $100 billion.
Thinking Machines achieved a valuation of $12 billion in its first funding round; SSI, established for less than a year, reached a valuation of $32 billion. These valuations reflect the scarcity of top AI teams while also indicating that a significant portion of future growth expectations has already been priced in.
The risk of the current AI primary market lies in this. If a few platform companies emerge in the future that have sufficient revenue and profit scales to support valuations in the hundreds of billions, then today’s high valuations may still be justifiable; but if most AI companies ultimately fail to establish sufficiently strong technological and business barriers, then excessively high entry prices will directly compress the potential returns for investment institutions.
Currently, this investment model still appears to perform well on paper. Data from Carta 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 significant portion of these gains comes from revaluation resulting from subsequent financing, rather than from actual cash returns generated by IPOs or acquisitions. The continuously rising primary market valuations today will ultimately need to be validated through future revenue growth and exit prices.
But for large VCs like Accel, the more pressing issue right now is not waiting for valuations to revert, but 如何避免错过这一轮技术周期中可能出现的少数赢家。
This is also one reason large funds continue to expand. The growth and financing cycles of AI companies are clearly shortening, and once a project gains market recognition, its valuation may rapidly increase over successive rounds of financing. If VCs want to secure sufficient equity early on and maintain ownership ratios during subsequent financing, they need to prepare capital reserves significantly higher than in the past.
Accel defines AI as a technology “super cycle” that is still in its early stages. Given this premise, the $3.5 billion early-stage and expansion fund and the previously raised $5 billion late-stage capital correspond to the same strategy: to enter potential winners as early as possible and reserve space for continued follow-on investments once valuations rise rapidly.
This also constitutes the most obvious contradiction in the current AI investment boom: AI has lowered some costs of entrepreneurship but has not decreased 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 price required to acquire and maintain a sufficient ownership stake in quality projects within an increasingly shorter window of opportunity.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
