Anthropic is set to price its initial public offering today, the last step before the AI company begins trading on Nasdaq under the ticker ANT tomorrow. The terms locked in for the roadshow call for 825 million shares priced between $70 and $78 each, split between roughly 801 million new shares sold by Anthropic itself and about 24 million sold by existing stockholders cashing out a portion of their stakes. At the top of that range, it is one of the largest technology listings in years; at the bottom, it is still large enough to reshape how public markets price generative AI.
The detail that has drawn the most attention from analysts is not the share count but who is standing behind it. Alongside the public offering, Anthropic has arranged a $26 billion private placement with seven strategic partners, and two of the commitments inside that placement are striking simply for who made them: Nvidia and Alphabet have each pledged $10 billion. Both companies sell Anthropic the infrastructure and cloud capacity its models run on, and both also build competing AI systems of their own, which makes a combined $20 billion vote of confidence from two rivals-cum-suppliers an unusually loaded signal for the rest of the market to read.
Exactly what that signal says about Anthropic's value depends entirely on which estimate you believe. Analyst commentary around the roadshow has ranged from a valuation near $1 trillion at the midpoint of the pricing range to figures exceeding $1.4 trillion once the private placement and strategic commitments are factored in. That spread, on the eve of an actual price being set, says less about Anthropic specifically than about how unresolved valuations remain across the entire AI sector, where revenue multiples that would be considered reckless in any other industry have become the market's working assumption.
Retail investors have been given an unusually generous slice of this one. Up to 20% of the offering has been set aside for retail allocation, a share far higher than most mega-cap technology listings typically offer individual investors, who are usually left bidding for scraps once institutional buyers have taken their fill. Whether that generosity reflects confidence in strong demand or a deliberate attempt to build broad public goodwill around a company whose products are now woven into everyday consumer and enterprise software, is a question the market's reaction over the following days will begin to answer.
Why suppliers investing in their own customer is unusual
Nvidia and Alphabet's $10 billion commitments each stand out precisely because of the layered relationship both companies have with Anthropic. Nvidia sells the GPUs Anthropic's models are trained and run on; Alphabet, through Google Cloud, sells Anthropic significant compute capacity while its own DeepMind division develops directly competing large language models. A supplier putting serious capital into a customer is not unheard of in technology, but it typically signals one of two things: genuine conviction that the customer's growth will keep driving demand for the supplier's own core business, or a more defensive calculation that locking in a strategic relationship is worth more than the capital itself, insurance against a rival chip or cloud provider winning that business instead. Markets watching Thursday's open will be parsing which motive investors believe is doing more of the work in a $20 billion combined bet from two companies with every reason to understand Anthropic's business model in granular detail.
The valuation gap nobody has closed
A spread this wide between analyst valuations, roughly $1 trillion to $1.4 trillion depending on how the private placement is weighted, is unusual even by the standards of a sector that has priced AI companies aggressively for several years running. Traditional valuation methods built around revenue multiples or discounted cash flow struggle with a company like Anthropic precisely because so much of its implied value rests on assumptions about AI adoption curves, enterprise spending growth and competitive positioning years into the future, assumptions analysts are visibly not converging on even hours before a price is actually set. That lack of convergence is itself a data point: it suggests the market has not yet developed a stable framework for pricing frontier AI labs the way it has for more established technology business models, leaving today's pricing decision to function as much as a price discovery exercise for the whole sector as a transaction specific to Anthropic.
What happens when trading opens tomorrow will matter well beyond Anthropic's own balance sheet. A listing of this size, backed by commitments from the companies that sell it chips and compete with it on models, sets a reference point for every AI company still weighing whether to go public, and for every investor trying to work out whether the sector's valuations reflect a genuine platform shift or a financing cycle that has simply not yet been tested by a market downturn.