$3.5 billion more for Anthropic

The AI landscape is evolving at a relentless speed—like an escalating game of 3D chess. Take Anthropic’s latest breakthrough: Claude 3.7, the first hybrid reasoning model, delivering both lightning-fast responses and deep, step-by-step problem-solving.

Alongside Claude 3.7 Sonnet, Anthropic unveiled Claude Code, an agentic coding tool in research preview. Seamlessly integrating into developers' workflows, Claude Code operates within the terminal, understands codebases, and accelerates development with natural language commands. Claude Code is designed to streamline and supercharge the coding process from automating routine tasks to managing version control, representing one of the most compelling examples of Agentic AI yet. 

These innovations cement Anthropic’s position at the forefront of AI, bridging practical problem-solving with cutting-edge technology. And in this high-stakes industry chess match, their latest moves are game-changing. 

That’s why we’re proud to back Anthropic again as part of their latest $3.5 billion financing.

Bessemer Venture Partners is doubling down on our investment in Anthropic because we are convinced that winning in AI requires more than engineering efficiency—it demands best-in-class talent, breakthrough research, massive hyperscaler capacity and distribution, and an unwavering commitment to safety.

Four key factors drove our decision to double down on Anthropic

1. Enterprise AI is at an inflection point: In 2023, the initial wave of AI adoption focused on consumer applications and proof-of-concepts. But in 2025, enterprises are moving from experimentation to production deployment. This shift demands more than just raw model performance—it requires robust infrastructure, reliable scaling, and enterprise-grade security. Anthropic demonstrates this enterprise-first approach, combining state-of-the-art capabilities with the features businesses need to deploy AI safely and effectively.

2. Developers are the key to platform dominance: Anthropic's commanding lead in code generation and software development capabilities represents a critical strategic advantage. Just as previous technology platform shifts like virtualization, cloud computing, and mobile were driven by developer adoption, AI's enterprise penetration has been led by its ability to enhance developer productivity through best-in-class models. Claude's superior performance in coding tasks, combined with its deep understanding of software architecture and best practices, has made it the preferred choice for developers worldwide. This developer-first approach isn't just about feature sets—it's about building a sustainable competitive advantage in what has historically been the most important early adopter user persona and use case in the enterprise. What makes this particularly powerful is the closed-loop nature of software development: unlike more subjective AI applications like content creation or writing, code either works or it doesn't. This immediate, objective feedback loop enables rapid iteration and improvement (via reinforcement learning), allowing Anthropic to build upon its existing significant lead in the coding use case while accelerating the path to enterprise-wide adoption.

3. The safety advantage is widening: Recent developments have only reinforced our belief that AI safety isn't just an ethical imperative—it's a crucial competitive advantage. While others rush to market with minimal safeguards, Anthropic's constitutional AI approach and comprehensive safety framework are becoming increasingly valuable to enterprises that can't afford the reputational or operational risks of unsafe AI deployment. Anthropic’s Responsible Scaling Policy has established an industry gold standard other competitors struggle to match. All of these activities are underscored by Anthropic’s continued commitment to research on mechanistic interpretability and alignment.

4. Distribution moats and growth matter more than ever: The partnerships Anthropic has forged with Amazon (Bedrock) and Google (Vertex) aren't just capital infusions, infrastructure access and distribution channels—they're strategic moats that become more valuable as the AI market matures. Customers prefer to leverage LLMs where their data already resides and where they already have commercial relationships . For many businesses, the ability to easily procure and integrate Claude within their hyperscaler’s environment is a real advantage. The early success of these partnerships, with Claude emerging as the most requested model on Amazon Bedrock, validates our thesis about the importance of enterprise distribution. And as we reflect on this reinvestment, what excites us most is Anthropic’s astonishing traction. Anthropic’s velocity of growth at 10-figure scale has only been matched by their current rival. As we’ve explored, the AI foundation model landscape will be one of the largest — and most competitive — markets in technology history.

Architects of the AI future

Recent AI efficiency developments are significant but represent just one aspect of this historic technical paradigm. The actual AI race to value isn't just about training costs—it's also about building AI systems that enterprises can trust, scale, and rely on. Anthropic's comprehensive approach to this challenge, combining cutting-edge research with practical enterprise needs, positions them uniquely in this race.

Dario and Daniela Amodei have cultivated an impressive concentration of AI talent at Anthropic and assembled a team that comprehends the immense potential and significant responsibilities of developing cutting-edge AI systems. The co-founders’ experience at OpenAI and the exceptional team they've assembled provide them with a distinct advantage in safely and successfully developing AI at scale.

This new influx of funding will accelerate Anthropic's research into more capable and safer AI systems, expand its enterprise infrastructure, and strengthen its position as a trusted partner for businesses navigating the AI revolution. As the market matures and enterprises demand more sophisticated and trusted AI solutions, Anthropic's comprehensive approach will prove increasingly valuable. The team’s multi-faceted vision to invest in improving underlying models, developing agentic frameworks, advancing computer use, and building upon their research in interpretability and alignment is incredibly compelling. As these all come together in the not-too-distant future, we anticipate that Anthropic will eventually provide virtual AI collaborators to knowledge workers and Enterprise customers of every shape and size. This isn't just incremental progress; it's a fundamental rethinking of how AI can serve human potential across every industry.

The next chapter of AI will be written by companies that can balance innovation with responsibility, technical excellence with practical deployment, and rapid progress with careful consideration of consequences. Anthropic possesses all these qualities—and more importantly, it has a team capable of executing this vision at scale.

If you're building AI infrastructure or the AI applications running on top of them, contact someone on our team. Learn more about our AI investment thesis and view our latest AI and data infrastructure market map.

Learn more about our journey with Anthropic:

• Why we backed Anthropic (October 2024)
• AI’s race to value with insights from Anthropic
• XB100: Our conversation on Deep Tech with Anthropic