Wall Street is selling off debt-heavy AI infrastructure companies, but private investors are moving in the opposite direction, pouring fresh capital into startups such as Etched, Unconventional AI and Legora. The two groups are also financed differently: the public companies carry heavy debt loads, while the startups profiled here have raised their capital through private equity rounds.

Etched Doubles Its Valuation in a Month

The most dramatic numbers belong to Etched, which builds chips designed specifically for AI inference—the stage at which a trained model actually generates answers.

In July 2026, Etched closed a Series C round of $300 million at a $10.3 billion valuation. One month later, the company raised another $700 million at a $21 billion valuation, a round that coincided with its first hardware delivery. A Series C sits late on the private funding ladder, and valuations do not usually double between successive rounds, which underlines how quickly money moved toward this company.

Investors are not paying purely for a pitch deck. Etched has received more than $1 billion in customer orders and has shipped its first inference system to Jane Street, the trading firm. The company has scaled its headcount to match, employing around 400 staff.

Unconventional AI and Legora Ride the Same Wave

Unconventional AI was founded in late 2025. In December 2025, it closed a $475 million seed round at a $4.5 billion valuation, ranking among the largest early-stage investments in AI hardware. A seed round is typically a company's first institutional raise, which makes a valuation of that size attached to it particularly notable. The company focuses on making computing more energy-efficient.

Legora, by contrast, sits on the software side. Founded in 2023, it builds AI tools aimed at the legal sector—an example of vertical AI, meaning software built around the specific workflows and compliance needs of one profession rather than a general-purpose product. In October 2025, Legora raised $150 million at a $1.8 billion valuation. In March and April 2026, it secured $550 million, later extended to $600 million, at a valuation of $5.55 billion to $5.6 billion.

Revenue figures help explain the enthusiasm. Legora serves more than 1,000 customers and reached $100 million in annual recurring revenue—a measure of the yearly value of a company's subscription contracts—within 18 months of launch.

Why Private Money Is Zigging While Wall Street Zags

The three startups attracting private capital share a few traits. Each targets a specific problem rather than trying to be everything to everyone: Etched is chasing inference, Unconventional AI is chasing energy efficiency, and Legora is chasing legal workflows.

Each also has some form of traction to point to. Etched has orders and a shipped system. Legora has customers and recurring revenue. Unconventional AI has a valuation that suggests investors are betting heavily on its team and technical approach.

All of the rounds above fall within roughly a year of each other, a window in which private investors were willing to fund these companies at sharply rising valuations while public markets sold off debt-heavy AI infrastructure names.

What This Means for the AI Investment Landscape

For chip incumbents, Etched's progress is worth watching closely. A specialized inference chip with over $1 billion in orders shows that buyers are willing to try alternatives for specific workloads. If more customers like Jane Street follow, general-purpose hardware could face pressure in the inference segment.

Unconventional AI's seed round highlights how central energy has become to the AI conversation, as power constraints increasingly act as a bottleneck for data centers.

Legora's trajectory points to vertical AI software as a serious category. Selling AI into a specific profession, with its own workflows and compliance needs, appears to produce faster revenue growth than broad tools.

The risks are real, though. Valuations that double in a month leave little room for stumbles. Etched now has to deliver hardware at scale, which is notoriously harder than shipping a first unit. Unconventional AI carries a multi-billion-dollar valuation at the seed stage, which raises expectations for every subsequent milestone. How each company handles those milestones—and whether private investors keep funding at these levels—will show whether the current divergence between public-market caution and private enthusiasm holds.