Benchmark violates its own guidelines by raising $2 billion and launching its inaugural growth fund.

      The discipline was integral to Benchmark's identity. For over twenty years, the firm maintained its funds at around $425 million, primarily investing in young companies with an approximate 20% stake in each, believing that selectivity rather than scale would yield returns. However, that approach has evolved.

      According to the Wall Street Journal, the firm has now closed $2 billion across two new funds, including its first growth fund of $1.25 billion focused on later-stage investments. Alongside this growth fund, Benchmark has also established a $750 million early-stage fund, which is larger than its historically raised funds and acknowledges that even seed and Series A investments need to be more substantial to remain competitive. The company, which had defined itself by its modest size, has recognized that maintaining such small operations has become limiting.

      The drawbacks of this constraint are apparent. Benchmark's smaller fund sizes likely precluded it from investing in capital-heavy AI laboratories, where funding rounds often exceed hundreds of millions; it did not invest in OpenAI, Anthropic, or the surge of foundation-model startups that have characterized the current cycle. Where it did venture into AI, the results were variable.

      Benchmark led a $75 million round in Manus, a Singapore-based agency platform that achieved $100 million in annual recurring revenue within eight months, only to watch Meta agree to acquire it for approximately $2 billion—until regulatory hurdles in China halted the deal in April, leaving Benchmark's stake uncertain.

      Recent transactions reflect this newfound flexibility. Historically, Benchmark invested at the Series A stage, but recently, it has backed two Series B companies: Gumloop, a no-code platform for creating enterprise AI agents, which raised $50 million from the firm, and Monaco, an AI-driven sales and CRM platform.

      Everett Randle, a general partner, has positioned this shift as a focus on relationships rather than funding stages, expressing to TechCrunch that the firm seeks deep connections with founders starting at any stage, be it seed, Series A, or Series B.

      The growth fund emerged from a significant success: Benchmark led Cerebras’s Series A in 2016 and subsequently raised a $225 million special-purpose vehicle to participate in a $1 billion pre-IPO round. Cerebras went public last month, yielding the firm $3.25 billion at the IPO price, a result that, according to someone familiar with the strategy, spurred the creation of the dedicated growth fund. This fund aims to make five or six large investments in both existing portfolio companies and new ventures.

      Changes extend beyond the funds themselves. Over two years, Benchmark has restructured its partnership, with Miles Grimshaw departing for Thrive Capital in 2024, Sarah Tavel shifting to a venture-partner role, and Victor Lazarte leaving to establish his own firm. In their stead, the firm welcomed Everett Randle from Kleiner Perkins and Jack Altman, brother of OpenAI’s Sam Altman.

      Collectively, these changes reveal a firm poised for a different landscape. Increased capital, extended investment stages, and new partners signal a transformative approach. Benchmark, which spent two decades asserting that restraint was its advantage, now acknowledges that in the AI era, such restraint has started to resemble a disadvantage.