That’s exactly what played out in internet search and ride-sharing, for instance. Few still remember Google’s early competitors. Lyft lags far behind Uber, which — despite years of skepticism — now looks set to generate $6B in free cash flow by 2025. Scale, network effects, and strong unit economics all converged to create a “winner takes it all” outcome.
Now the spotlight shifts to foundational models. OpenAI recently raised $40B (subject to contingencies) at a $300B valuation—a private IPO of sorts. Market leaders aren’t cheap… ChatGPT has surpassed 20M paid subscribers and 500M weekly active users, and is projected to generate $13B in 2025, despite a $5B loss.
Will OpenAI be another “winner takes it all” case, even with heavyweight rivals like Grok, Llama, and Gemini in the race?
Fabrica Ventures believes this won’t be the case, which is why we recently invested in Anthropic (Claude model).
Claude rivals ChatGPT in performance — with new benchmarks emerging every month, it’s not worth diving into all the bells and whistles. But most importantly, Anthropic was designed with alignment, security, and steerability at its core — essential features for developers and enterprise adoption.
Foundational models monetize through two primary channels: consumer applications and enterprise APIs. While OpenAI clearly dominates the consumer space, Anthropic is quickly establishing itself as the strongest name in the enterprise market.
Despite OpenAI’s scale and network effects, we don’t see enterprise adoption as a winner-takes-all scenario. Ethical concerns surrounding OpenAI only reinforce this view. And in practice, many enterprises are likely to deploy multiple models, both for redundancy and to maintain leverage in negotiations.
Anthropic scaled revenue from $10M to $1.0B in just two years and is projecting $2.5B for 2025 — with 85% of it driven by enterprise API usage.
Anthropic has become the top choice for AI coding. Moreover, its offering extends well beyond a “traditional” GenAI LLM, covering AI agent development, autonomous agents & digital coworkers, personal AI assistants & co-pilots.
Now valued at $61.5B and backed by $11.5B in funding — 70% of it from Amazon and Google (even as Google develops its own Gemini model) — Anthropic is extremely well-capitalized to be one of the winners among developers and enterprises.
Conclusion
In ABBA’s mega-hit The Winner Takes It All, the “loser stands small”.
But in foundational models, we believe the outcome will mirror the cloud market, where AWS, Azure, and Google Cloud successfully coexist.
In our view, Anthropic has the potential to become a landmark company, one of the most iconic success stories in VC history.
Post Scriptum
Last week, Anthropic rolled out an API for AI-powered web search. Perhaps the winner takes it all … until a new techno paradigm shift flips the game.