Key Takeaways
- • Bessemer Venture Partners' Sameer Dholakia said the firm's Anthropic investment at a $15 billion pre-money ($18 billion post-money) valuation was the lowest-voted deal in the firm's history, and it has since become Bessemer's largest growth-fund position at roughly 20% of one vehicle, according to his appearance on the 20VC podcast.
- • Dholakia prices companies on next-twelve-month (NTM) revenue, saying he will pay 10 to 15 times NTM for fast-growing businesses, and he cites an implied current-ARR multiple of about 150x on Anthropic's Series D entry at roughly $100M ARR.
- • Bessemer raised $5.75 billion across a $1.75 billion seed and early-stage fund and roughly $4 billion across early and late growth funds, managed by the same four general partners.
- • Dholakia argues venture is bifurcating into global platforms and extreme specialists, warning that firms in the middle of the market face a difficult position.
- • Dholakia predicts Waymo could become a trillion-dollar business by expanding into personal car ownership, trucking and logistics, and says Bessemer invested nearly $200 million in Waymo at a $115 billion valuation.
Summary
Bessemer Venture Partners growth investor Sameer Dholakia said on the 20VC podcast that the firm backed Anthropic's Series D at a $15 billion pre-money, $18 billion post-money valuation despite it being the lowest-voted deal in Bessemer's history. Two partners overrode the room, and the position has since grown to roughly 20% of one growth vehicle. Dholakia's rationale rests on a "cloud redux" model: frontier AI is so capital-intensive that it consolidates to a few players, much as cloud consolidated to AWS, Azure and Google Cloud, while many applications are built on top. He said he regretted not writing a bigger check in that round, and noted that AWS's investment proved pivotal for training compute and distribution.
Dholakia says he prices companies on forward twelve-month revenue rather than trailing ARR. He cited an implied current-ARR multiple of about 150x for Anthropic, about 90x for Legora, and 20x or more for Field Guide, and said he would pay 10 to 15 times NTM for businesses growing faster than he has seen in 31 years. He also said he passed on OpenAI at a $30 billion valuation after ChatGPT, having imagined only a 3x return.
On open versus closed models, Dholakia argued that the precise mix of tokens matters less than the size of the addressable market, which he framed as intelligence rather than oil. He said most enterprises will prefer a provider to run open-source models, and he sees no bear case for frontier labs. Bessemer's sovereign-AI investments include Sarvam in India and Mistral in France. On competition, he said capital is necessary but not sufficient, and that the old T2D3 growth benchmark no longer attracts capital in an environment where companies can grow 10x at scale.
Bessemer's $5.75 billion in total commitments comprise a $1.75 billion seed and early-stage fund and about $4 billion across early and late growth. The firm targets 5% to 10% of a growth fund per company, but will go to 20% for generational companies. Dholakia described venture as bifurcating between global platforms and extreme specialists, with mid-sized funds at risk. He said Waymo's addressable market extends well beyond ride-hailing, calling it a potential trillion-dollar business, and he expects AI-driven drug discovery to compress a half-century of progress into a decade.
Dholakia says he prices companies on forward twelve-month revenue rather than trailing ARR. He cited an implied current-ARR multiple of about 150x for Anthropic, about 90x for Legora, and 20x or more for Field Guide, and said he would pay 10 to 15 times NTM for businesses growing faster than he has seen in 31 years. He also said he passed on OpenAI at a $30 billion valuation after ChatGPT, having imagined only a 3x return.
On open versus closed models, Dholakia argued that the precise mix of tokens matters less than the size of the addressable market, which he framed as intelligence rather than oil. He said most enterprises will prefer a provider to run open-source models, and he sees no bear case for frontier labs. Bessemer's sovereign-AI investments include Sarvam in India and Mistral in France. On competition, he said capital is necessary but not sufficient, and that the old T2D3 growth benchmark no longer attracts capital in an environment where companies can grow 10x at scale.
Bessemer's $5.75 billion in total commitments comprise a $1.75 billion seed and early-stage fund and about $4 billion across early and late growth. The firm targets 5% to 10% of a growth fund per company, but will go to 20% for generational companies. Dholakia described venture as bifurcating between global platforms and extreme specialists, with mid-sized funds at risk. He said Waymo's addressable market extends well beyond ride-hailing, calling it a potential trillion-dollar business, and he expects AI-driven drug discovery to compress a half-century of progress into a decade.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.