Cherubic Ventures Closes $68.88M Fund VI for AI

Cherubic Ventures closed Fund VI at $68.88M on September 1, 2026, bringing firm-wide assets under management above $500M across six funds. The Taipei-founded venture firm will use the vehicle to back early-stage AI-native companies across infrastructure, developer tools, enterprise software, healthcare, physical AI, and robotics.

The repeated 8s in the fund size are deliberate. In East Asian culture, 8 is associated with prosperity and good fortune. The memorable figure sits on top of a more serious investment question: how does a first-check fund keep finding consequential AI companies before larger pools of capital and market consensus arrive?

Cherubic Ventures is led by Founder and Solo GP Matt Cheng. The firm says companies in Fund VI have already raised more than $500M in subsequent financing. That number belongs to the portfolio companies, not Cherubic's own AUM, but it helps explain the model. Cherubic is trying to enter before the evidence is obvious, then let a broader financing market carry the companies through later rounds.

What Cherubic Ventures Announced

The Fund VI announcement describes a final close of $68.88M, not a target, launch, first close, or interim close. With the vehicle complete, Cherubic reports more than $500M in AUM across its six funds.

Cherubic did not identify individual Fund VI limited partners. The firm said investors across all six vehicles include global institutions and foundations, public companies, family offices, entrepreneurs, and high-net-worth individuals. It also did not disclose Fund VI's check-size range, ownership targets, reserve ratio, investment period, or audited performance.

The accounting deserves a clean line. Fund VI is $68.88M. Cherubic's firm-wide AUM is above $500M. Fund VI portfolio companies have raised more than $500M from subsequent investors. Those three figures describe different pools of capital, and combining them would turn a useful early-stage story into venture arithmetic by mood.

Why a Smaller First Check Still Matters in AI

Artificial intelligence has expanded the amount of capital some startups can absorb, especially in robotics, physical infrastructure, model development, and compute-intensive software. That can make the biggest vehicle look like the most important one. Early-stage investing works on a different clock.

A pre-product or pre-market-fit company rarely needs one investor to finance its entire future. It needs an investor willing to decide while the product, buyer, and category are still unsettled. Cherubic's stated edge is making that first institutional commitment across the United States and Asia, then supporting founders as the companies become legible to larger funds.

That is why the more than $500M of subsequent financing inside the Fund VI portfolio is interesting without being treated as a return metric. The figure indicates that later investors have already entered companies Cherubic selected early. It does not reveal Cherubic's ownership, realized gains, marks, or fund-level performance.

Fund V closed at $110M in 2022, when the firm reported more than $400M in AUM. Fund VI is smaller by headline amount. Cherubic has not said the difference reflects a deliberate contraction, a different reserve policy, or a revised ownership strategy, so the public evidence does not support inventing one.

The AI Portfolio Behind Fund VI

Cherubic says Fund VI is focused on AI-native companies across infrastructure, developer tools, enterprise software, healthcare, physical AI, and robotics. The named portfolio illustrates how broad that mandate can become.

Sudo AI is the most prominent example in the announcement. The robotics company was co-founded by embodied-AI researcher Hao Su and entrepreneur Robin Han. Cherubic says it was Sudo AI's earliest institutional investor and reports that the company reached a valuation near $2B two years after its founding.

Sudo AI's robotic system is trained through virtual simulation and is designed to handle unfamiliar objects without requiring real-world manipulation data for every new task. That claim matters because physical AI companies have to move beyond carefully prepared demonstrations and perform amid the variation of actual environments. The near-$2B valuation is company-reported, however, and should not be confused with an audited operating result or a realized Cherubic return.

Cherubic also identifies Entire, the developer platform founded by former GitHub CEO Thomas Dohmke. Entire raised $60M earlier in 2026, according to the firm. Other Fund VI names include AI patent platform Patlytics, healthcare-operations company Max AI, longevity company Generation Lab, and drug-development company therapiAI.

The portfolio spans different customers and capital needs, but the investment logic is consistent. Cherubic is searching for founders before the market has settled on the product language, distribution model, or category winner.

The Solo-GP Decision Model

Cherubic describes itself as one of the earliest venture firms to adopt the solo-GP model. Matt Cheng holds the final investment identity, while the firm's current team includes partners Danielle Dudum, Pierre Arys, Tina Cheng, and Snow Hua across investment, operations, portfolio management, and investor relations.

That structure concentrates judgment. It can make an early decision faster and keep the firm's thesis coherent, but it also makes selection quality unusually visible. There is less room to blame an investment committee when a market moves in a different direction.

Cherubic's historical portfolio gives the model credibility. The firm invested early in companies including Hims & Hers, Flexport, Calm, Paidy, 91APP, and Astranis. Hims & Hers and 91APP became public companies, while PayPal acquired Paidy for $2.7B. Those outcomes belong to earlier vehicles and do not prove Fund VI performance, but they explain why limited partners would continue backing an approach built around pre-consensus entry.

What Fund VI Changes

Fund VI gives Cherubic another $68.88M to compete for the first meaningful position in companies that may require far larger financing later. The fund's value will depend less on matching the scale of billion-dollar AI vehicles and more on reaching founders before those vehicles know where to look.

The disclosed portfolio already shows the handoff in motion. Cherubic supplies the first institutional conviction; later investors bring larger checks as products, customers, and categories become easier to underwrite. That relationship is becoming more important as AI companies combine software economics with the capital demands of laboratories, healthcare deployment, robotics hardware, and compute infrastructure.

For founders, the practical question is whether Cherubic's global network and early-stage judgment can remain useful after the first check. For limited partners, the question is whether concentrated solo-GP selection can keep producing access that a larger process cannot manufacture on demand. Fund VI has closed. The companies it found before consensus now have to show what that head start was worth.