Bessemer's $5.75 Billion Fund Bets Early and Growth-Stage Capital Belong Together
Bessemer's $5.75B fund splits early-stage and growth investing into one vehicle. Here's what it signals about how fast founders expect to scale.
Bessemer Venture Partners closed $5.75 billion in new capital on September 23 split into two pools: $1.75 billion for seed and early-stage companies and $4 billion for growth-stage bets. The structure is not new, but the scale and timing reveal something important about where large venture firms expect the next wave of returns.
By bundling early and growth capital in a single fund rather than raising separate vehicles, Bessemer signals confidence that it can identify winners at inception and that those winners will need capital fast. The model assumes AI-native companies will compress the timeline between first funding and scale-up rounds, making it essential for leading investors to hold both buckets of capital ready.
The single fund with two stages
Bessemer's approach differs from how venture capital structured funds for decades. Traditionally, firms raised separate funds for different stages: a seed fund, an early-stage fund, a growth fund. Each served a distinct investor base with different return expectations and holding periods. Founders shopped around, and firms that backed a company at seed often competed with other investors to lead subsequent rounds.
The dual allocation within one fund simplifies this. The same limited partners fund the full journey. The same investment team deploys both buckets. Byron Deeter, a Bessemer partner, noted that "AI-native companies are scaling faster than any category of technology we've backed before." That speed prompted the expanded growth practice and the consolidated fund structure.
Bessemer has already deployed $3 billion across more than 260 AI-native companies since 2022. The portfolio spans the AI stack: compute infrastructure, foundation models, developer tools, applications and agents. The size of the new fund and its allocation suggest the firm expects that pace to continue or accelerate.
Why capital concentration by stage is eroding
The pressure on fund structures comes from startup timelines. AI startups now raise capital earlier and at younger ages than their non-AI peers, with a median age at first financing 65 percent lower.
This compression squeezes the intermediate stages. If a company needs seed money at 18 months old and growth capital at 36 months, the traditional model of separate-fund timelines breaks. A growth fund raised in year two won't reach its investment period until years three through eight. By then, the companies that needed that capital in year three have either secured it elsewhere or failed.
The traditional model also created friction for successful companies. A founder who impressed a seed investor faced competition to lead the Series A, but had no guarantee that seed backer would win. The new model changes the calculus: a founder backed by Bessemer at seed knows Bessemer can supply all follow-on capital if the company performs. That continuity reduces search costs for founders and gives Bessemer clarity about its own capital deployment.
A broader shift in VC structure
Bessemer is not alone.
Other large firms like Andreessen Horowitz run multiple funds across stages and sectors specifically so founders can stay with the firm throughout growth cycles. The pitch to limited partners emphasizes partner continuity: the same person who invested at seed sits on the board at Series B and C. That consistency of perspective and network effects compounds over time.
The trend accelerated in 2026 as capital concentrated around AI. Global venture investment reached $300 billion in the first quarter alone, with AI companies capturing a disproportionate share. That concentration meant a small number of firms controlled the capital founders needed most. A single closed round at Bessemer or Andreessen Horowitz could shift a company's trajectory. Firms that controlled both early and growth capital had an advantage: they could shape the entire arc of a company's development.
“AI-native companies are scaling faster than any category of technology we've backed before.”
What the structure signals about returns
The Bessemer fund allocation—$1.75 billion early, $4 billion growth—tells us something about expected returns and risk. Venture returns concentrate in growth-stage outcomes. Companies that reach scale can return the entire fund to investors. Seed checks rarely do. The 30-70 split (early to growth) shows Bessemer's confidence that its early picks will graduate to growth rounds, and that growth rounds will be the primary value driver.
This allocation pattern also reflects a shift in where venture capital thinks the competition matters most. If every firm can fund seed rounds—and they can, given capital abundance—the edge comes from the ability to spot companies early, back them through explosive scaling, and exit at high valuations. Early-stage capital is table stakes. Growth capital is where returns concentrate.
By consolidating both in one fund, Bessemer also manages concentration risk differently. A dedicated growth fund that underperformed would need to raise a smaller follow-up fund or shut down. A single fund that combines both stages can redeploy capital between buckets if early bets scale slower or if growth opportunities disappear. The structure is more flexible.
The implicit bet on continuity
The structure assumes that Bessemer will be able to identify winners early enough to fund them through scale-up. That is not guaranteed. Many seed-stage bets fail before needing growth capital. Others succeed but founder preferences, board conflicts, or better offers from other investors lead companies to seek different growth backers.
The model also assumes venture returns will remain concentrated in the hands of a few firms. If capital becomes more abundant, or if new competitors emerge with specific sector expertise, Bessemer's advantage of early identification erodes. Founders backed at seed by Bessemer might still shop Series A rounds if other investors offer better terms, networks, or stage-specific expertise.
Bessemer's deployment history suggests it has earned the track record to back this strategy. The portfolio includes Anthropic, Perplexity, and other AI companies that have scaled rapidly. Past success, however, does not guarantee future returns in a market moving as fast as AI venture capital.
