How much has Lambda raised?

Lambda has raised over $3.1 billion in disclosed capital since its founding in 2012, spanning equity rounds from seed through Series E and three significant debt or credit facilities. The equity component totals roughly $2.4 billion across seven rounds; the remainder is structured debt — a $500M GPU-backed credit facility (April 2024), a $275M credit facility (August 2025), and a $9.5M debt facility alongside the 2021 Series A — each secured against GPU hardware assets or revenue contracts.

The capital structure reflects the asset-heavy nature of GPU cloud infrastructure: Lambda must purchase or lease NVIDIA hardware before it can sell compute, creating a 'procure compute, resell at margin' business that demands substantial upfront capital. Gross margins of approximately 50% overall (61% excluding legacy hardware lines) suggest the model works at scale, but the company ran a net loss of roughly $24 million in H1 2025 as it reinvested revenue into capacity growth ahead of the IPO.

A notable structural element: the NVIDIA $1.5 billion, four-year leaseback deal announced in September 2025 is technically a customer revenue contract, not a funding round, but it functions like one — locking in $375 million per year in predictable, take-or-pay revenue from Lambda's largest chip supplier. That single contract materially de-risks the Lambda investment thesis for public market investors.

Lambda's cap table spans four distinct investor cohorts. First, AI-specialist early-stage investors: Gradient Ventures (Google's AI fund), 1517 Fund, Bloomberg Beta, and Georges Harik backed Lambda from Series A through B, betting on AI infrastructure before the ChatGPT demand wave. Mercato Partners led the Series B, citing Lambda's infrastructure positioning in the growing AI training market.

Second, growth equity and strategic funds: USIT (Thomas Tull's US Innovative Technology Fund) led the Series C at Lambda's unicorn milestone, then co-invested in the Series E. Andra Capital and SGW co-led the Series D. Individual AI luminaries — Andrej Karpathy (ex-OpenAI Director of AI, ex-Tesla AI Director), Greg Brockman (OpenAI co-founder), Garry Tan (Y Combinator President), and Adam D'Angelo (Quora CEO) — participated across rounds, lending credibility to Lambda's positioning within the AI research community.

Third, strategic industrials: NVIDIA invested in the Series D alongside hardware ODMs Supermicro, Pegatron, Wistron, and Wiwynn — each with a direct commercial interest in Lambda as a demand channel for their GPU hardware. In-Q-Tel, the CIA's investment arm, also joined the Series D, signaling defense/government interest. Fourth, sovereign and megafund capital: TWG Global — backed by Mubadala Capital of Abu Dhabi — led the Series E and the subsequent convertible note round, providing the sovereign-scale capital needed for gigawatt-class infrastructure.

Lambda's valuation trajectory — from $87.5M in 2021 to $5.9B in late 2025 (a 67x increase in four years) — tracks almost perfectly with the explosion in demand for NVIDIA GPU compute following the launch of ChatGPT in late 2022. Each funding round was timed to either precede or immediately follow a major demand inflection. The Series B in March 2023 came just months after ChatGPT ignited enterprise AI spending; the Series C in February 2024 followed 2023 revenues of approximately $250M (a dramatic step-up from prior years); and the Series E followed the multibillion-dollar Microsoft deal that provided years of revenue visibility.

Secondary market data from Forge Global suggested Lambda's implied valuation on private secondary trades reached as high as $11–$12 billion in early 2026 — well above the $5.9B Series E price — a signal that institutional investors expect a significant re-rating at IPO. The $350M convertible note discount (20% below the expected IPO price) implies that the company and its bankers see a reasonable IPO floor above $5.9B.

For context, CoreWeave — Lambda's closest comparable — completed its IPO in March 2025 at a $14.2B valuation and had grown to a ~$64B market cap by June 2026, a 350% gain in roughly 15 months. If public markets assign comparable revenue multiples to Lambda, a significantly higher IPO valuation than the $5.9B Series E price is plausible, which explains the secondary market premium and the strategic logic of the convertible note's 20% discount structure.

Lambda is not yet profitable on a net-income basis. H1 2025 net loss was approximately $24 million on revenues exceeding $250 million, suggesting the company could approach breakeven as revenues scale — annualized revenue reached approximately $760 million by late 2025. Gross margins of ~50% are healthy for a capital-intensive infrastructure business, but Lambda is deliberately reinvesting into data center buildout rather than optimizing for near-term earnings.

Lambda has explicitly targeted an H2 2026 IPO. The $350M convertible notes carry a one-year IPO deadline with financial penalties for non-completion, creating strong structural pressure to list. Investment banks Morgan Stanley, J.P. Morgan, and Citi have been formally retained. Competitor CoreWeave completed a successful public offering in March 2025, validating public market appetite for AI infrastructure companies — and CoreWeave's ~$64B market cap as of June 2026 provides a visible comparable that makes Lambda's IPO valuation case straightforward to model.

New CEO Michel Combes — who has led public companies including Sprint and Alcatel-Lucent — was hired specifically to execute the IPO and the transition to gigawatt-scale operations. The parallel appointment of a new CFO (Charles Fisher, ex-Turo) and CLO (David Connolly, ex-Altice) is establishing the financial controls and governance structures required for a public company.

Lambda's capital position signals two things for vendors and sellers. First, buying power is substantial and actively deployed: a company sitting on $3.1B in cumulative capital and generating $760M+ in annualized revenue can make large procurement decisions quickly. Lambda is actively expanding its data center footprint — Chicago AI factory (23MW+ with EdgeConneX), Kansas City AI factory (100MW+ planned, 24MW initially), and a 2GW+ long-term buildout vision — meaning active procurement cycles for power infrastructure, cooling systems, networking equipment, high-density rack solutions, and storage.

Second, organizational maturity is accelerating fast. The May 2026 leadership overhaul — new CEO, CFO, CLO, and Chairman — signals a company professionalizing its procurement, finance, and legal functions ahead of an IPO. Vendors who engage now, while new executives are still establishing vendor relationships and defining preferred-vendor rosters, have an advantage over those who wait until post-IPO procurement formalization locks in incumbent relationships.

The one caution: Lambda's build-not-buy posture is strong for anything in the critical GPU compute path. The engineering team builds its own orchestration, monitoring agents, scheduling layers, and much of the ML toolchain. The higher-opportunity vendor categories are adjacent: financial systems (ERP, FP&A, procurement), legal and compliance (CLM, e-signature, entity management), security (SOC 2 compliance tooling, network security for InfiniBand fabrics), and high-performance distributed storage at scale.