The IPO Deal in One Look

The figure that has settled is $1.4 trillion. That works out to roughly $70 to $78 per share with 825 million shares outstanding. The first day of trading is expected to be November 19. The Nasdaq ticker is set to be ANT, and the company is looking to raise somewhere around $60 billion. Twenty percent of the shares are set aside for regular retail buyers.

Chip and cloud companies are buying $26 billion of the stock at the IPO price. Google and Nvidia each get $10 billion. Bloomberg has reported that the offering is multiple times oversubscribed. Big buyers are forcing the price upward, and many want in early to capture discounts before the stock reaches the open market. There is concern that this setup could lead to a sharp drop once trading begins, with the possibility of heavy short selling. Early investors may see this as a chance to cash out before any crash. That is the prediction being made around the current dynamics.

Two Anthropics on the Road

There are two versions of Anthropic circulating this week. One appears in the legal filings, where lawyers check every sentence because statements can lead to lawsuits. The other lives in the roadshow video, broker calls, and Friday’s live stream with Jamie Dimon. That version carries the big numbers. The distinction matters for anyone trying to understand what is being presented versus what can be scrutinized.

Revenue Targets and Dario’s February Warning

Back in February, Dario said that if revenue is not $1 trillion—even if it reaches $800 billion—there is no hedge on earth that could stop bankruptcy if the company buys that much compute. Analysts have since been told that the next quarter will slow down significantly. Traders have still bid the valuation up to $1.8 trillion anyway.

The average employee is sitting on around $41 million in stock. Some of them can begin to sell on day one. That raises a direct question: are buyers purchasing the company that reaches $1 trillion in revenue, or the one where employees with tens of millions in stock start walking out the door? The compute commitment is described as so far off the chart that it is crazy. A Chamath post quoting the situation and adding eye emojis has drawn over 160,000 views.

Current Numbers and the Path to 2028

Revenue in Q3 was $17 billion. Operating losses that quarter were almost half a billion dollars—in fact over half a billion, nearly $600 million. This is not a profitable company. Dario has said that if the business does not reach a trillion, the math does not make sense.

In the same quarter, OpenAI reported $10 billion in revenue and lost almost $11 billion. Counting the stock it pays people, the loss figure is about $13 billion. Anthropic’s losses already include stock compensation; OpenAI’s reported numbers do not. Bloomberg has noted that the two companies count revenue a little differently, using different accounting approaches.

In August, Reuters reported that Anthropic’s own forecast for all of 2028 was $190 billion to $200 billion. A $1 trillion run rate at the end of 2028 would mean roughly $250 billion in a single quarter. From the last quarter’s $17 billion, that requires a 10x to 15x increase in revenue. That kind of growth is not seen as realistic in the available time. Right now the company has over $900 billion committed in CapEx. It is unclear whether any organization has spent that much. At best, Q4 is projected to reach $19 billion in gross revenue. That remains a far cry from the growth needed within two years.

SpaceX, Data Centers, and the Compute Build-Out

There is talk that Anthropic will destroy SpaceX in their IPO. That may be possible, but SpaceX is currently receiving a large amount of revenue from Anthropic. The company has said it plans to end that relationship by the end of next year. After recent elections, however, building data centers faces a significant uphill battle. It will likely be much harder for Anthropic to construct the compute capacity it expects. Getting off the SpaceX supply by next year is not foreseen. Very few can build data centers at the scale Elon can right now, so the need for that capacity is expected to continue.

Local Compute as the Alternative Path

Local compute is expected to keep catching up. Teams building on OpenMonoAgent and similar harnesses are positioned to take market share from the large cloud-dependent models. The work currently underway in Open Mono Agent does not require the huge cloud providers. In many cases a $20 or $30 per month subscription is used to generate plans that are then fed to local AI for execution. Frontier models can handle web search, image work, and similar tasks to produce a detailed plan, which is then handed to local compute to run. In plenty of situations even that step is unnecessary.

This approach supports custom software solutions for organizations that want their own stack, data sovereignty, and to avoid sending customer data into the cloud. Startuphakk brings more than a decade of executive experience as a fractional CTO and 25 years in software development to transform tech teams and products, including custom AI solutions. OpenMonoAgent.ai allows an entire stack to be set up with one command so teams can start building custom AI software themselves.

Most companies do not have a technology problem; they have a leadership problem that shows up in missed deadlines. Bad technology decisions waste money and can kill companies. The businesses succeeding right now are not simply chasing the latest AI trends. They build on solid engineering and treat AI as infrastructure they own, control, and integrate into software that actually works. That is the model delivered by the custom software team at StartupHakk: database architecture, API design, system integration, and scalable infrastructure built the way it has been done for over 25 years. When AI belongs in a solution, it is designed into the architecture and runs in the customer’s environment, not a vendor’s cloud and not on someone else’s pricing schedule.

OpenMonoAgent.ai is the open-source platform that demonstrates the approach: a terminal-native AI coding agent running entirely on local LLMs, with zero API costs, zero telemetry, and full ownership. It is growing because engineers recognize real infrastructure. As a fractional CTO engagement, the same standard of strategic architecture, hands-on delivery, and clear accountability is available without the full-time executive cost.

The upcoming Anthropic IPO rests on a $1.4 trillion valuation, a $1 trillion revenue target by the end of 2028, and compute commitments already exceeding $900 billion. Current quarterly revenue of $17 billion, substantial operating losses, and the gap between internal 2028 forecasts and the broker pitch leave the math under scrutiny. Whether the market prices the ambitious roadshow numbers or the practical constraints around compute, employee stock, and growth remains to be seen once trading begins on November 19. For those who prefer to own and control their AI rather than rent it from frontier providers, local stacks and tools such as OpenMonoAgent.ai offer a direct alternative.