AI Startup Valuations Soar Even as Big Tech Stocks Wobble

A Trillion-Dollar Question

Wall Street got a jolt recently when the so-called ‘Magnificent Seven’ — Apple, Microsoft, Amazon, Alphabet, Meta, Nvidia, and Tesla — collectively shed roughly $797 billion in market value in a matter of days 1. For a group of companies whose stock gains have been closely tied to the artificial intelligence narrative, the sudden reversal has reignited a familiar question: is the AI trade overheated, or even a bubble on the verge of popping? Adding to the unease, IBM's most recent quarterly results showed guidance cuts and signs that its AI ambitions are lagging behind its steadier hybrid cloud business, even as the company leans on dividend appeal to reassure investors 5.

Private Markets Tell a Different Story

Yet while public megacap tech stocks wobble, the private funding market for AI startups shows no such hesitation. Fireworks, an AI infrastructure company backed by Nvidia, just reached a $17.5 billion valuation as it pivots away from heavy reliance on coding-startup client Cursor and toward a broader base of companies seeking cheaper, more efficient AI models 2. That shift — toward lower-cost inference and model efficiency — echoes a wider industry trend of enterprises trying to control ballooning AI compute bills.

Elsewhere, unicorns are being minted across very different corners of the AI ecosystem. GlossGenius, the beauty-industry software company founded by twin sisters, raised $44 million at a $1.15 billion valuation and rebranded as Genius AI to expand beyond salons and spas into small-business tools more broadly 3. Reid Hoffman, along with Mark Pincus and Ritankar Das, is meanwhile negotiating a $100 million raise for their AI research lab Prentis at a projected $1 billion valuation, adding another well-connected founder team to the growing list of AI unicorns 4.

Chasing Nvidia's Crown

Hardware is drawing outsized bets too. Sequoia-backed Etched raised $300 million in a Series C round at a $10.3 billion valuation, positioning itself as one of a growing cohort of AI chip startups trying to chip away at Nvidia's dominance 6. TensorWave, a Las Vegas-based cloud infrastructure startup explicitly marketing itself as an Nvidia alternative, secured a $1.55 billion valuation in its latest round, betting that data-center customers want more diversified chip supply chains 7.

Reading the Divergence

Taken together, these developments suggest a split-screen moment for AI: public megacap valuations are facing scrutiny and volatility, with IBM's stumble underscoring that even established players can disappoint on AI execution, while venture capital continues pouring billions into startups spanning chips, infrastructure, and vertical software. Whether that private-market enthusiasm is a leading indicator of durable value creation or simply lagging the same speculative fervor now rattling public markets remains the open question hanging over the entire sector.