Close to 40 VC-backed startups reached billion-dollar valuations in the first half of 2026, according to a TechCrunch compilation published on July 5, 2026, drawing on Crunchbase and PitchBook data. The monthly pace—31 new companies in January, 37 in March, 29 in May—marks the strongest half-year volume since the 2021 funding peak.

AI is the dominant driver. Manufacturing and security each contributed three new unicorns in January, both sectors explicitly propelled by AI integration. Fintech placed fourth overall, generating ten billion-dollar startups, including Juspay, an Indian payments infrastructure company.

The Standout Raises

Several May entries deserve closer attention for their scale and positioning within the current infrastructure buildout.

Recursive, an AI research lab founded in 2025, closed a $650 million Series A led by GV and Greycroft, reaching a $4.65 billion valuation. That valuation-to-age ratio—roughly $4.65 billion after one year—would have been unrealistic in previous funding cycles. MiRus, a medical device company, hit $4.41 billion after a $1.5 billion late-stage strategic round from Boston Scientific, the largest single check in the cohort.

EXA, which builds web retrieval infrastructure for AI agents (think of it as systems that help AI programs search and retrieve information from the web), raised $250 million in a Series C led by Andreessen Horowitz, landing at $1.95 billion. Positron, an AI hardware startup founded in 2024, reached $1.06 billion on $234 million in Series B funding, signaling continued investor appetite for alternatives in the inference hardware space—the chips and systems that run AI models after they're trained.

On the healthcare side, Vi Labs closed a $145 million round at $1.64 billion, and Forus, a healthcare automation company, raised $160 million in a Series B led by Accel to hit $1.01 billion. The two illustrate the breadth of "AI in healthcare": clinical workflow automation alongside enterprise platform products.

Socket, a software supply-chain security startup, raised $60 million in a Series C led by Thrive Capital to reach $1 billion. Supply-chain security has been a persistent focus since the SolarWinds and Log4Shell incidents; Socket's milestone reflects maturing enterprise procurement decisions rather than speculative enthusiasm. SendCutSend, which produces custom industrial parts using digital fabrication, raised $110 million in a Series A led by Paradigm and Sequoia, signaling broadening investor interest in physical-world automation. Inventory management platform Radar raised $170 million in a Series B to reach $1 billion, while wealth management platform Farther closed a $150 million Series D led by General Atlantic at $1.25 billion. In June 2026, MainFunc—the company behind AI workspace Genspark—completed a $485 million Series B, bringing total funding to $645 million and pushing its valuation to $2.6 billion.

What the Pattern Suggests

The sectoral spread here is wider than the 2021 vintage, which was heavily concentrated in fintech and SaaS. This cohort spans inference hardware, industrial fabrication, supply-chain security, and wealth management. AI either directly defines the product or sits beneath the hood as an operational accelerant in the majority of cases.

The timing and pace of these valuations merit attention. Recursive at $4.65 billion on a single Series A one year after founding, and Positron at $1.06 billion two years in, reflect a market where investors are pricing expected infrastructure dominance rather than demonstrated revenue scale. Early infrastructure cycles have historically rewarded first-mover commitment—but that logic places significant weight on technical execution and go-to-market speed that few young organizations have yet proven they can sustain at scale. It's a bet on execution, not on current business metrics.

Geographically, the May cohort tilted toward established ecosystems: 17 of 29 companies were US-based, with the UK and China each accounting for four. This concentration is consistent with a broader re-centering of early-stage AI capital around regions with deep technical talent and existing networks, even as the underlying models and tools grow more accessible globally. With the pace running at roughly six to eight new unicorns per month, the 2026 full-year total is tracking toward the highest count since 2021.