Anthropic (Claude) IPO - Worth Buying?
Video: Reviewing Anthropic's business fundamentals as they gear up for an IPO.
Transcript
Anthropic, the artificial intelligence company behind the Claude AI models, is planning to IPO this year. There’s no set date yet, but estimates are sometime in Q3 or Q4 of 2026. They might raise as much as $60 billion when they go to market.
The company was founded by seven former OpenAI researchers, led by the siblings Dario Amodei and Daniela Amodei. They had disagreements with OpenAI on safety and commercialization, thus Anthropic was born.
As of the most recent funding round in February, Anthropic is currently valued at $380 billion in private markets. Investors included Amazon, Google, Microsoft, Nvidia, and a variety of equity and sovereign wealth funds.
AI is obviously still a hot industry, and institutional and corporate investors want in just as much as retail will when this thing finally goes public.
Business Model
Diving into their business model, if you know anything about AI, their main path to revenue is by selling subscriptions:
70%-75% of revenues come from API consumption, usage in corporate/developer environments with customers paying-per-token for model access.
10%–15% of revenues comes from Direct-to-Consumer subscriptions (Claude Pro/Max)
And another 10%–15% of revenues come from specific enterprise partnerships
Keep in mind, Claude has only recently been in the news after their falling out with President Trump over safety guidelines Anthropic was not willing to budge on. Most consumers had no idea what Claude was until then. This company has been focused on B2B, with 70%+ of its revenue being generated from enterprise customers.
OpenAI’s ChatGPT is, of course, the AI model everyone tends to think of when they think of AI since it became a phenomenon in 2022.
Anthropic has over 300,000 business customers and 70% of Fortune 100 companies are Claude customers.
The business model is obviously capex-intensive. Training frontier AI models costs billions per run. They have to build costly data centers. Anthropic has committed to roughly $80 billion in cloud infrastructure investments, buying more cloud services and computing power from big tech firms through 2029.
They have also committed to investing $50 billion to build AI data centers across the United States, including facilities in Texas and New York.
The company is unprofitable and burned approximately $3 billion in cash in 2025 and does not expect to reach cash-flow breakeven until 2028. Gross margins are currently around 50%, with management projecting as high as 77% margins by 2028 as they scale and renegotiate compute contracts.
Their revenues are a mixture of recurring subscriptions and usage-based consumption. Subscriptions give them some revenue predictability, but the majority of their revenue is from usage, which can be volatile. The enterprise focus does provide stickier revenue dynamics than a more consumer-focused model like ChatGPT.
Growth Opportunities
Anthropic has proven it can consistently release new features for consumers. Their current best model, Claude Opus 4.6, has consistently ranked as the top model in a competitive industry. Granted, these rankings change all the time, as these AI firms are constantly releasing new models.
The recent releases of Claude Code, a command-line coding agent, and Claude Cowork, their daily-task productivity agent… have impressed consumers in the space. In less than a year, Claude Code hit $2.5 billion in annualized revenue. Business subscriptions for the service have quadrupled in the first couple of months in 2026.
Also, something that just happened recently, we saw Anthropic make the mistake of leaking the code for their Claude Code application. Not the code for their actual AI models, but the Claude Coding service.
The issue was caused by human error, when someone over at Anthropic left a source map file, used for debugging, in a public registry that allowed developers to reconstruct the source code. Massive blunder by the company. No customer data was leaked, but throwing your code out for everyone to see does not help you maintain a lead over the competition. But I digress…
Claude is also the only AI model available across all three major cloud platforms. AWS Bedrock, Google Cloud, and Microsoft Azure. No other model developer has that level of distribution right now.
Anthropic is estimating that global AI spending will hit $2 trillion in 2026, and the company is positioned to capture a disproportionate share of that through its enterprise sales focus.
Anthropic’s management team has been projecting:
$20 billion to $26 billion in revenue for 2026.
Upwards of $55 billion in 2027.
Upwards of $148 billion by 2029.
It’s difficult to know what products they might release to increase revenues further, the industry is constantly evolving. But even just current growth rates with their various subscriptions and model usage is more than enough on its own. Everyone knows AI is seeing insane growth.
Anthropic is still in the early innings of penetrating European, Middle Eastern, and Southeast Asian enterprise markets. Any region, really.
The company’s current annual recurring revenue (ARR) as of this month… is running at approximately $19 billion, according to Bloomberg. That is up from $9 billion at the end of 2025. So, their ARR is rising rapidly as Claude’s popularity has grown. Up 111% in just three months…
If you assume this $19 billion run rate is going to stick, or grow, then Anthropic is currently trading at nearly 20x revenues. Given their latest funding round, which valued them at $380 billion.
This valuation assumes they’re going to continue running incredible growth rates. And it’s hard to see how they won’t. Who knows what market cap the IPO might be placed at, but no doubt it will be pricing in perfection. Any slip-up or market crash, and investors will get punished.
Competitors
A key aspect of investing in the AI industry and Anthropic, is going to be… who wins the AI race? The competitive landscape has 4 primary competitors remaining.
OpenAI is the most direct competitor. With a projected $29.4 billion in full-year 2026 revenue and 800 million weekly users, using ChatGPT. OpenAI has a larger consumer business and higher overall revenue.
However, OpenAI is burning cash at a far higher rate… projecting $74 billion in operating losses through 2028. They are not expected to reach profitability until 2030, two years after Anthropic’s target.
OpenAI was recently valued at $850 billion in private markets. The key competitive difference between these two is that Anthropic leads in enterprise adoption. Anthropic has roughly 40% market share versus OpenAI’s 25%, while OpenAI dominates the consumer side of the market.
Google DeepMind is the second major competitor, with their Gemini models. Google does have the advantage of massive existing assets in cloud and search infrastructure. Google holds roughly 21% enterprise AI market share. Google has also invested $3 billion in Anthropic and provides TPU access, so the relationship is both competitive and cooperative.
Meta is number three. Meta does not sell model access via API but has become the primary force behind open-source AI through its LLaMA models. Meta’s strategy is to commoditize the base layer of AI, aka the model itself, and make AI access cheap or free.
Instead, the value would then shift to who owns the applications built on top of the open source models. This would directly undermine the business case for paid API access from companies like Anthropic.
If open-source, free models can do most of what Claude can do over time through open development… then Anthropic has to do more to justify their subscriptions, right.
xAI is the last competitor worth mentioning, via Grok. xAI has raised significant capital and recently merged with SpaceX ahead of its own expected IPO. It has less enterprise traction but benefits from Musk’s distribution network through Twitter.
There are a few Chinese players, some in other regions, but it’s clear that the American tech companies are the ones with all of the best models right now.
Who gives up, out of these four companies, could cede billions of dollars in revenue to the other players… so it will be interesting to see if any other companies drop out of the AI race.
Meta is in their own world, but it’s possible that we see either OpenAI or xAI have to drop off because of the massive capex required to compete.
Granted, it seems like these companies have endless levels of capital to access from private equity and corporate investors.
So, it’s hard to see any of them drop out unless liquidity dries up.
Competitive Moat
Moving on, we have to consider how much of a moat Anthropic really has
In theory, barriers to entry help them avoid new competition entering the industry by now, investing billions in building new data centers or computing power is simply not viable for most companies.
With that said, they are in a cutthroat competition with the remaining tech companies. So, that’s not much of an advantage.
Switching costs can be a pain for enterprise and retail consumers, once you’re using a model in all of your workflows… it’s a hassle to switch everything over. But that can be done. And the U.S. government is. They are swapping over from Claude to ChatGPT.
Overall, the industry is firmly in oligopoly territory, so the AI market is currently a tit-for-tat between the top companies on who has the best model. And the sector is constantly evolving over time.
As consumers, that competition works in our favor, but as investors, it leads to a brutal competitive landscape.
Risks
Diving into the risks of investing in Anthropic, I think it should be pretty obvious…
Capital-intensive data centers are burning billions in investor capital. All of these companies are running large losses. The AI arms race requires continuous massive investment in compute, and any slowdown in revenue growth could force Anthropic to raise more capital.
As we just went over, competition is fierce from both open-source and walled-off AI models, granted the industry is finally starting to consolidate.
And then the primary risk is probably some form of major government crackdown on Anthropic specifically or the AI sector at large. Potential copyright infrigement issues are still unsolved. We know Trump wants an AI legislation moritorium, but anti-AI sentiment has been growing… so. That’s something to consider.
In summary, you might struggle to find any other companies in the world that can grow as quickly as the AI model developers. Every big tech company you can think of is all in on this industry, besides Apple lol.
Investors will undoubtedly love what they’re seeing. I’m sure many retail investors are foaming at the mouth to get a chance to buy some of these pure-play AI companies. But the risks are certainly there, and you could see some black swan absolutely blow these companies to smithereens after they IPO.
With that said, I am under the assumption that they will reach profitability in the future. Much like Uber has relatively recently. It just took the industry quite some time to mature to that point. Anthropic is targeting 2028, and it wouldn’t surprise me at all to see them reach profitability.
I don’t tend to invest in stocks this large regardless, but especially not when the global stock markets and financial economy are in a dire situation thanks to the war in the Middle East. Either way, Anthropic should be dropping their S-1 form relatively soon, which would give us more info on their operations. So, investors will have to wait for that and see what they think.
Disclaimer
This content is for informational purposes only, you should not construe any such information or other material as legal, tax, investment, financial, or other advice. Nothing contained in this video constitutes a solicitation, recommendation, endorsement, or offer by Green Investing to buy or sell any securities or other financial instruments in any jurisdiction.
All content in this video is information of a general nature and does not address the circumstances of any particular individual or entity. You alone assume the sole responsibility of evaluating the merits and risks associated with the use of any information or other content in this video before making any decisions based on such information.