Anthropic Launches IPO Roadshow, Targets Up to $1.4 Trillion Valuation
The AI software market watches Anthropic's public debut as labs release new models optimized for specialized agentic and efficiency tasks.
The story
Anthropic officially launched the roadshow for its initial public offering this week, aiming for a valuation of up to $1.4 trillion. The company plans to sell 825 million shares between $70 and $78 each, potentially raising over $61 billion.
This move sets a significant public market benchmark for frontier AI labs, following earlier private funding rounds that valued the company at $965 billion. The IPO comes as Anthropic reported $33.2 billion in revenue for the first nine months of 2026, with an annualized run rate reaching $81 billion by October. The public offering is a key test of investor confidence in the AI sector amid broader market volatility, especially as questions persist about the sustainability of massive AI investments.
Who moved
Anthropic
What Changed: The company launched its IPO roadshow, targeting a valuation of up to $1.4 trillion and aiming to raise over $61 billion by selling 825 million shares.
Consequence: This move establishes a public market valuation benchmark for frontier AI labs, providing significant capital for accelerated expansion.
Google DeepMind
What Changed: Gemini 4 Carbon, a high-performance model, was made available to developers, enterprises, scientists, and consumers.
Consequence: This expands access to Google's most capable model for demanding tasks like cybersecurity, complex coding, and scientific research.
Meta
What Changed: The lab released Muse Ember, a small and efficient model designed for complex reasoning, tool calls, and long document parsing.
Consequence: Muse Ember offers developers a cost-effective solution for building fast, low-cost agentic workflows, matching performance of higher-priced models at a fraction of the cost.
Cohere
What Changed: CEO Aidan Gomez indicated no firm timetable for a public offering, despite the company's revenue doubling annually for the past three years.
Consequence: This signals a cautious approach to public markets, contrasting with other AI labs and focusing on continued private growth with government backing.
Astral
What Changed: The AI lab raised $5.5 million in funding to expand its hyper-personalized AI tutor to families across all 50 US states.
Consequence: This investment will support the broader deployment of its educational AI, which aims to help students master grades five times faster.
New models
Claude Fable 5.2
Lab: Anthropic
What: This model offers improved speed, lower pricing, and enhanced performance in coding, reasoning, and data-collection tasks.
Use: It is designed for general-purpose applications and excels in areas like formalizing mathematical proofs and complex problem-solving.
Claude Mythos 5.2
Lab: Anthropic
What: A configuration of Fable 5.2 with cyber and biology safeguards, offering advanced threat discovery and vulnerability scanning capabilities.
Use: Available to verified organizations through Anthropic's Cyber Verification Program for security and defensive reasoning tasks.
Muse Ember
Lab: Meta
What: Meta's smallest and most efficient model, built for complex reasoning execution, tool calls, coding subtasks, and long document parsing.
Use: It is intended for developers seeking a fast, low-cost model for agentic workflows and for teams aiming to reduce costs by running it as a subagent.
Gemini 4 Carbon
Lab: Google DeepMind
What: Google's strongest model with frontier performance in cybersecurity capabilities and long-horizon agentic workflows across knowledge work, software engineering, and scientific reasoning.
Use: Available via API for developers and enterprises tackling complex, multi-step coding, debugging, and open-ended scientific research.
Market signals
Anthropic launched its IPO roadshow, targeting a valuation of up to $1.4 trillion and planning to raise up to $64.4 billion.
Implication: This move establishes a significant public market valuation for a frontier AI lab, providing a key test of investor appetite in the sector.
SoftBank reported a $7 billion fiscal Q2 net loss, yet its stake in OpenAI reached a fair value of $99.6 billion, yielding $45 billion in paper gains.
Implication: This illustrates the mixed financial landscape in the broader tech market, where substantial AI investments are generating significant unrealized gains despite operational losses elsewhere.
The Canadian government announced a C$1.6 billion investment in Cohere.
Implication: This signals national support for domestic AI companies and a focus on securing AI capabilities for governments and regulated industries.
Australia announced the creation of a dedicated AI safety regulator with powers to levy significant penalties.
Implication: This indicates increasing governmental scrutiny and a proactive approach to managing AI risks, potentially setting a precedent for other nations.
Builders saying
Jack Dorsey, Block CEO. He announced Buzz 1.0 as a flexible, model-agnostic team workspace, emphasizing the need for adaptable AI environments that integrate various models for collaborative work.
What we'll be watching
- Anthropic shares are set to begin trading on the Nasdaq, with pricing on November 18 and trading commencing November 19.
- Microsoft Ignite conference takes place from November 17-20 in San Francisco.
- The AWS re:Invent conference is scheduled for November 30 to December 4 in Las Vegas.
- The NeurIPS academic conference will be held from December 6-12, with the main venue in Sydney.
- Watch for official dates and details regarding Google I/O and Microsoft Build 2027, typically announced in early next year.
Reporting + analyst voices: grounded via Google Search at publish time.