Venture Capital Funding Shifts to the Application Layer

Reach Capital closed a $265 million fifth fund on Tuesday, the largest in the San Francisco firm’s eleven-year history. The money will go into AI companies working across learning, health, and work, at checks of $1 million to $10 million from pre-seed through Series A.

The size of the fund is not the story. The story is where a specialist investor decided to point it, because venture capital funding is drifting away from model builders and toward the companies that sell something on top of them.

Inside the Fund V Numbers

Reach plans to back roughly 50 companies over the next three years, with about a quarter of the capital held back for follow-on rounds. The fund brings the firm close to $1 billion under management. Its earlier vehicles were smaller, and the step-up shows the pace institutional money is now moving at.

Past bets include Replit, ClassDojo, and Coral Care, which spans developer tooling, classroom software, and pediatric care. That range is deliberate.

From Edtech Specialist to Category Agnostic

Reach spent a decade known as an education investor. Education technology funding cooled hard after its pandemic peak, and rather than shrink alongside the category, the firm widened the aperture.

Generative AI is what made that possible. A company teaching workers a new skill now looks like education software, enterprise software, and recruiting technology at once, so the old sector labels stopped sorting deals usefully.

The firm frames its thesis around AI that expands “human potential,” a phrase partners used in announcing the fund. Translated into investor terms, it means backing applications rather than infrastructure. Founders who have watched AI startup funding concentrate at the top of the stack should recognize the counter-move.

What Application Layer Means for Your Pitch

Model costs keep falling. That is bad news for anyone whose pitch is a thin wrapper, because the underlying capability gets cheaper for your competitors at exactly the same rate it gets cheaper for you.

It is good news for anyone with a real workflow, proprietary data, or a customer relationship that is annoying to replace. Investors are underwriting durability now, not novelty, and durability comes from the parts of your business that are not the model.

So expect the diligence questions to shift. Instead of asking which model you use, partners will ask what happens to your margin when inference gets 80% cheaper, and whether your customers would notice if you swapped providers overnight.

How to Position a Raise in This Market

Lead with the workflow, not the technology. Show the specific job your product finishes and the measurable time or money it saves, because that is the claim a partner can defend in an investment committee.

Bring evidence of retention. Revenue that renews without a sales push does more for a valuation right now than a large logo that churns quietly in month seven.

• Name the workflow you own end to end.
• Show gross margin at current and projected model prices.
• Quantify switching cost for your best customers.
• Explain what data you accumulate that a rival cannot buy.

And size the raise to the milestone, not the market. Founders who watched a record seed funding round make headlines this summer should remember that a larger round mostly raises the bar for the next one.

The Fund Cycle Signal to Read Next

Watch for other specialist firms making the same move. If health, fintech, and industrial funds start describing themselves as AI application investors, capital availability at seed will improve for founders outside the model-building race.

Watch inference pricing too, since it sets the floor under every application company’s margin. Our reporting on rising AI spending tracks how quickly those costs move.

Public data helps here as well. The National Venture Capital Association research library publishes quarterly deal and fundraising figures that show whether a single fund close reflects a broader pattern.

Common Questions About Raising From AI Funds

Does a specialist fund still want non-AI companies?

Increasingly, no. Most now expect AI to be part of how the product works, not a feature added later.

Is pre-seed money getting easier to raise?

Selectively. Funds like this one write small first checks, but they concentrate them in a handful of workflows they already understand.

How much should follow-on reserves matter to me?

A great deal. A fund holding a quarter of its capital for follow-ons is far more useful in a hard second round than one that is fully deployed.

The takeaway for founders is short: the money is moving to whoever turns cheap intelligence into something a customer will pay for twice.