How Anthropic Grew from a $124 Million Startup to a Reported IPO Filing

Anthropic’s path from an early-stage AI company to a reported IPO candidate has been marked by rapid fundraising, the growth of its Claude products and increasingly prominent legal and policy disputes. The company raised $124 million in its 2021 Series A; by 2026, reports said it had confidentially filed for a U.S. IPO. That filing is not a completed listing, and reported revenue run rates and future forecasts should not be confused with revenue already earned.

Key takeaways

• Anthropic’s funding history includes a $124 million Series A in 2021, a $580 million Series B in 2022, and reported later valuations that rose sharply.
• Claude became the company’s main product family. In 2026, Anthropic separated broad access to safeguarded Claude Fable 5 from invitation-only access to Mythos 5. 11
• The IPO remains prospective. A reported $47 billion annualized revenue run rate and a $190–$200 billion 2028 revenue projection are different kinds of figures: the first extrapolates a recent pace, while the second is a forecast.

From early funding to strategic investment

Anthropic, led by siblings Dario and Daniela Amodei, announced a $124 million Series A in May 2021. It raised $580 million in a Series B in April 2022.

Claude arrived in 2023 as Anthropic’s AI assistant. That year, Amazon announced plans to invest up to $4 billion, and Google agreed to invest up to $2 billion. Those commitments were significant sources of strategic backing, but they should be distinguished from the company’s named funding rounds.

In March 2025, Anthropic raised $3.5 billion at a reported post-money valuation of $61.5 billion. A further $13 billion round in September 2025 was reported at a $183 billion post-money valuation. Claude Code, introduced as a standalone product in February 2025, added a coding-focused offering to the Claude family.

Reports placed Anthropic’s valuation at $965 billion in May 2026. That is a reported private-company valuation, not an IPO price or a guarantee of what public-market investors would pay.

Claude and Mythos: wider access versus restricted access

Anthropic continued expanding Claude’s capabilities and products, but the company’s 2026 Mythos announcements also highlighted its cautious approach to access. It introduced Claude Mythos Preview in April through a limited initiative. In June, Anthropic announced Claude Fable 5 as a safeguarded model for general use, alongside Claude Mythos 5, which it offered to a small group of cyberdefenders and infrastructure providers through Project Glasswing. 11

Anthropic describes Mythos as particularly capable in cybersecurity and biology research. Mythos 5.1, announced in September, remained limited to a small set of vetted organizations. These are the company’s descriptions of its models; they should be read as capability claims, not independent proof that a model will perform safely or reliably in every setting. 2

Copyright cases: a major settlement and a separate music lawsuit

In July 2026, a federal judge approved a $1.5 billion settlement resolving authors’ class-action claims over books used to train Claude. The settlement resolves those claims; it is not a universal ruling on whether all uses of copyrighted material to train AI models are lawful. Earlier proceedings had distinguished legally acquired books from pirated copies. 20

A separate case was filed in August by music publishing businesses associated with Sony and Warner Music. Their complaint alleged that Anthropic misused copyrighted song compositions to train Claude. The reporting provided here describes the lawsuit’s filing, not a final outcome. 17

Dispute with the U.S. government over military contracts

Anthropic’s conflict with the U.S. government centered on safety restrictions for its technology and the Pentagon’s exclusion of the company from military contracts. On September 25, 2026, a federal appeals court ruled 2–1 against blocking that exclusion. The decision marked a setback for Anthropic’s challenge, not a resolution of every question about how AI systems should be used by the military. 18

IPO status and the financial outlook

Anthropic reportedly filed confidentially for a U.S. IPO in June 2026. A confidential filing is a step in the listing process, not confirmation that an IPO will go ahead or when shares might begin trading.

Reuters reported that Anthropic’s annualized revenue run rate had exceeded $47 billion by May 2026 and that the company projected revenue of roughly $190 billion to $200 billion in 2028. The first figure extrapolates a recent pace of business; the second is an internal forecast, not a reported result. Separate September reporting said Anthropic’s annualized run rate could exceed $100 billion during 2026, another forward-looking estimate rather than a full year of realized revenue.

The distinction matters when assessing the company’s progress: fundraising and private valuations measure investor commitments and expectations, while a run rate and a multiyear forecast describe different ways of projecting revenue. Until a public offering is completed and more financial information is disclosed, Anthropic’s IPO timeline, valuation and future growth remain uncertain.