AI Valuations Soar as Anthropic Eyes Near-Trillion IPO

What happened

The AI funding frenzy shows no sign of slowing, even as fresh cracks appear in the broader tech market. The headline event is Anthropic's confidential filing for an IPO on July 25, 2026, reportedly targeting a valuation near $965 billion — a figure that would place the AI lab among the most valuable companies ever to go public 1. That single number captures the scale of investor conviction in generative AI, but it arrives alongside a wave of smaller, still-substantial funding rounds that show money flowing into nearly every layer of the AI stack, from chips to infrastructure to vertical software.

Nvidia-backed inference startup Fireworks has reached a $17.5 billion valuation, a business that pivoted away from heavy reliance on coding startup Cursor toward serving a broader set of companies looking for cheaper AI models 2. On the chip side, Sequoia-backed Etched raised $300 million at a $10.3 billion valuation, positioning itself as one of several challengers trying to chip away at Nvidia's dominance in AI silicon 7. In infrastructure, Las Vegas-based TensorWave — explicitly framed as an anti-Nvidia bet in cloud computing — secured a $1.55 billion valuation in its latest round 8.

Further down the valuation ladder, GlossGenius, the beauty-industry software company founded by twin sisters, raised $44 million at a $1.15 billion valuation while rebranding as Genius AI to expand into small-business tools beyond beauty 3. Prentis, an AI research lab co-founded by LinkedIn's Reid Hoffman, Zynga's Mark Pincus, and Ritankar Das, is separately negotiating a $100 million raise that would value it at $1 billion 4.

Set against this run of unicorn-minting deals is a starkly different narrative: a reported single-day rout in which the so-called Magnificent Seven — Apple, Microsoft, Amazon, Alphabet, Meta, Nvidia, and Tesla — collectively lost $797 billion in market value 5. Meanwhile, IBM's latest quarterly results reportedly showed a guidance cut, with hybrid cloud strength failing to offset lagging AI momentum, prompting analysts to reassess the company's valuation and dividend appeal 6.

Where the reporting agrees

Across the board, these accounts agree that capital is still pouring into AI at eye-popping valuations, whether for foundation-model leaders, chip challengers, or infrastructure upstarts 1278. There's also consistency in the theme that competition is intensifying beneath Nvidia and the largest labs — Fireworks, Etched, and TensorWave are each described as responding to demand for alternatives to the dominant, more expensive incumbents 278. And multiple pieces converge on the idea that the AI investment boom and signs of market strain are unfolding simultaneously, not sequentially — big valuations and big selloffs are being reported in the same window 156.

Where it doesn't

The most significant divergence is in framing rather than fact. The Anthropic IPO story and the Magnificent Seven selloff story both come from the same outlet, thetechedvocate.org, but they tell almost opposite tales: one treats near-trillion-dollar AI valuations as a triumphant milestone 1, the other treats a comparable dollar figure — $797 billion erased from seven of the world's biggest companies — as evidence the market may be overheating 5. Neither piece reconciles the two, leaving readers to square a record-setting private valuation with a historic public-market drawdown happening in roughly the same period.

There's also a gap in sourcing and confirmation. Anthropic's $965 billion figure is described as a reported valuation tied to a confidential filing, not a confirmed, priced outcome 1, which matters because confidential IPO filings routinely see their eventual valuations shift before shares actually trade. The IBM piece, by contrast, is framed around specific financial results — a guidance cut and a quarterly miss — presented as company-reported fact rather than speculation 6. That's a meaningfully different evidentiary footing than the unconfirmed Anthropic number.

Finally, no single source attempts to connect the unicorn-funding wave to the Magnificent Seven selloff or IBM's stumble. Each data point — Fireworks, Etched, TensorWave, GlossGenius, Prentis — is reported as an isolated deal 23478, not as part of an explicit bubble narrative, even though the cumulative picture looks like one.

The reading that holds up

Taken together, the evidence supports a market that is bifurcating rather than uniformly bursting. Private AI funding — for chips, inference, infrastructure, and vertical software — remains robust and largely insulated from public-market jitters 23478. But the reported Magnificent Seven selloff and IBM's AI-lag guidance cut suggest public investors are growing more discriminating about which AI bets justify their valuations 56. The Anthropic IPO filing is the real test case: if a nearly trillion-dollar valuation survives public scrutiny, it would validate the private market's exuberance; if it doesn't, the coming months could reveal just how much of the AI boom's valuation is grounded in current revenue versus future promise.