Understand any stock in simple English
Enter a ticker to learn what the company does and how it makes money.
Anthropic builds AI models that companies and people pay to use for writing, coding, and answering questions.
Anthropic is an artificial intelligence company. It builds large language models — software trained on enormous amounts of text that can read a question, reason through it, and write back an answer, a piece of code, or a whole document. People and businesses use its models, sold under the Claude name, to draft emails, summarize long reports, write and fix computer code, and answer questions in plain language.
The money comes from selling access to those models. Developers pay per word processed through an API — the on-ramp that lets other software plug into Claude — while individuals and teams pay a subscription for the chat app. Big companies pay for seats and for extra security and controls. Anthropic also earns from cloud partnerships: its models are rented out through Amazon Web Services and Alphabet's Google Cloud, so customers who already run their systems on those clouds can switch Claude on without moving their data. What makes Anthropic unusual is its structure and its pitch. It is a public benefit corporation, meaning its charter commits it to balancing profit with its stated mission of building AI safely, and it was founded in 2021 by former OpenAI researchers, including siblings Dario and Daniela Amodei. That safety-first framing is also a sales tool: banks, hospitals, and government agencies that are nervous about AI want a vendor they can defend to regulators. Its main rivals are OpenAI and Alphabet's Gemini, and the competition is fierce on both capability and price.
Dario Amodei walked away from OpenAI at the end of 2020, after fights over how fast to sell AI and how safe to make it. In early 2021 he started his own lab with six other former OpenAI researchers, including his sister Daniela, and they built the company as a Public Benefit Corporation — a legal structure meant to let it chase AI safety and the public good, not just shareholder money. A $124 million first round led by Skype co-founder Jaan Tallinn gave the new lab room to hire and think. The problem they were solving was simple: the people building the most powerful AI wanted a place where caution came first, not last.
In early 2023 Anthropic shipped Claude, its first big language model — a direct shot at OpenAI's ChatGPT. Rather than chase consumers, it handed the model to enterprise partners like Notion, Quora, and DuckDuckGo for writing, summarizing, and coding. That was the moment the safety lab became a commercial AI provider.
A few months later, Amazon said it would put up to $4 billion into Anthropic, making AWS its main training partner and cloud provider; Google followed with up to $2 billion. It was cash, but it was also a leash: Anthropic's computing now ran on the giants' clouds, and that shaped its costs for years.
Early in 2025 Anthropic launched Claude Code, an AI tool built right into a developer's daily workflow. Instead of fighting for consumer chatbot traffic, it sold engineering productivity to companies willing to pay. Revenue exploded from about $4.6 billion in 2025 to over $16 billion in the first half of 2026, with nearly 12,000 enterprise customers spending more than $100,000 a year each.
In early 2026 the Trump administration moved to restrict Anthropic's government business after the company refused to support mass domestic surveillance and autonomous weapons. The Pentagon declared Anthropic a supply-chain risk, and Anthropic sued to block the label that spring. A court largely sided with Anthropic that summer, but an appeals court later upheld a separate supply-chain risk designation — a reminder that safety promises can cost you your biggest customer.
In mid-2026 Anthropic quietly filed to go public, then made its paperwork public in late October, planning to list on Nasdaq under the ticker ANT. It was aiming for a valuation of $1.5 trillion to $2 trillion — which would make it the largest IPO in history, bigger than SpaceX's offering earlier that year. The filing showed first-half revenue of $16.3 billion, up about 1,300 percent, alongside a net loss of $80.7 billion, mostly from non-cash accounting charges.
The business is growing fast but is expensive to run. Training and serving giant models requires fleets of specialized chips and huge data centers, so Anthropic leans heavily on cloud partners for computing power and on outside investors — Amazon and Google have poured billions into the company — to fund the buildout. Revenue has climbed sharply as enterprise adoption took off, but the company has been burning cash to keep pace. That is the bet: spend enormously now to win the customers and safety reputation that make the business durable later.
Saved answer · last researched 27 Oct 2026