Lambda Customer Base, Marquee Accounts, and Fleet Scale
| Cohort / Account | Figure | As of | Note |
|---|---|---|---|
| AI developers and teams (total) | 200,000+ | Oct 2025 | Lambda: 'currently serves over 200,000 AI developers and teams worldwide' |
| Customers (count) | Tens of thousands | Nov 2025 | Lambda's own description in its Series E release |
| Microsoft | Multibillion-dollar, multi-year | Nov 2025 | Tens of thousands of NVIDIA GPUs including GB300 NVL72 systems; Lambda has worked with Microsoft for more than eight years; exact terms undisclosed |
| Kansas City single-customer deployment | >10,000 NVIDIA Blackwell Ultra GPUs | Planned early 2026 | Dedicated to a single Lambda customer under a multi-year agreement; footprint expected to double over time |
| Named customer stories | SPREEAI, Kodiak, Pika, Iambic Therapeutics, fal, Meshy, Genesis Therapeutics | Nov 2026 | Listed on Lambda's customer stories page |
| Lambda Stack users | Used by the F500, research labs, and the Department of Defense | Nov 2026 | Lambda's description of its software stack on its product page |
| Supercluster scale (per deployment) | 4,000 to 165,000+ GPUs | Nov 2026 | Single-tenant, shared-nothing AI cloud |
| 1-Click Cluster scale (per deployment) | 16 to 2,000+ GPUs | Nov 2026 | Self-serve InfiniBand-connected clusters for 2 weeks to 1 year |
Lambda Data-Center Power Capacity Buildout (MW toward 3 GW)
| Site / Milestone | Status (date) | Power Capacity (MW) | Detail |
|---|---|---|---|
| Prime Data Centers LAX01, Vernon CA | Leased (Nov 2025) | 21 MW initial (33 MW campus) | NVIDIA Blackwell GPU infrastructure in Vernon's first AI-ready data center |
| Kansas City, MO AI factory | Planned to open early 2026 (announced Oct 2025) | 24 MW initial; potential 100+ MW | >10,000 NVIDIA Blackwell Ultra GPUs at launch, expected to double over time; Lambda is sole tenant |
| EdgeConneX, Chicago and Atlanta | Announced Aug 2025; Chicago due for service in 2026 | 30+ MW (23 MW Chicago) | Hybrid liquid-to-chip and air cooling; single-tenant build-to-density site in Chicago |
| 2030 capacity target | Target (2030) | 3,000 MW (3 GW) | AI compute under management by 2030, the goal Lambda's leadership team is built to reach |
| 2030 GPU target | Target (end of decade) | n/a (over 1,000,000 GPUs) | EdgeConneX says Lambda's 2GW+ vision leads to over a million GPUs deployed by the end of the decade |
Lambda Funding Rounds, Valuation, and GPU Debt
| Round / Instrument (date, lead) | Amount Raised | Valuation / Terms | Notable Investors / Lenders |
|---|---|---|---|
| Series A + debt facility (2021-07) | $24.5M ($15M Series A equity + $9.5M debt) | Valuation not disclosed | 1517, Gradient Ventures, Bloomberg Beta, Razer, Georges Harik; Silicon Valley Bank (debt) |
| Series B (2023-03, Mercato Partners) | $44M | Valuation not disclosed | Mercato Partners (lead), Adam D'Angelo, Jeff Hammerbacher, Greg Brockman, Lukas Biewald, Garry Tan, 1517, Gradient Ventures, Georges Harik, Bloomberg Beta, Crescent Cove |
| Series C (2024-02, US Innovative Technology Fund) | $320M | Valuation not disclosed | USIT (lead), B Capital, SK Telecom, T. Rowe Price Associates, Crescent Cove, Mercato Partners, 1517 Fund, Bloomberg Beta, Gradient Ventures |
| Series D (2025-02, Andra Capital and SGW) | $480M | Valuation not disclosed | Andra Capital and SGW (co-leads), Andrej Karpathy, ARK Invest, Fincadia Advisors, G Squared, In-Q-Tel, KHK & Partners, NVIDIA; Pegatron, Supermicro, Wistron, Wiwynn; 1517, Crescent Cove, USIT |
| Series E (2025-11, TWG Global) | Over $1.5B | Valuation not disclosed | TWG Global (lead), USIT, existing investors |
| Senior secured credit facility (2025-08, upsized 2026-05) | $275M, upsized to $1B | Syndicated, multi-tranche; secured | J.P. Morgan (lead arranger) |
| Senior secured term loan B (2026-08) | $926M | SOFR + 3.00%, issued at 99.5%; matures 31 Dec 2030; Moody's Baa2 | Morgan Stanley (lead left arranger, bookrunner), MUFG (joint bookrunner) |
| Delayed-draw term loan (2026-10) | $1.008B | 6.78% fixed; matures 30 May 2033; A (low) Morningstar DBRS, Baa1 Moody's | J.P. Morgan (sole coordinating lead arranger, structuring agent, bookrunner) |

Lambda Labs in AI Software
KPI data tables
Lambda Labs operating metrics behind the KPI charts, newest period first (up to 24 periods per chart). Periods use each company's own fiscal labels where the source does; each table names its source.
| Period | On-demand list price (USD/GPU-hour) |
|---|---|
| B200 SXM6 | 6.99 |
| H100 SXM | 4.29 |
| GH200 | 2.29 |
| A100 SXM 40GB | 1.99 |
| A10 | 1.29 |
| V100 | 0.79 |
Source: lambda.ai/pricing (instances pricing, 1-GPU tier, as of 2 Nov 2026, https://lambda.ai/pricing): V100 $0.79 (8-GPU instance only), A10 $1.29, A100 SXM 40GB $1.99, GH200 $2.29, H100 SXM $4.29, B200 SXM6 $6.99; 8-GPU HGX B200 $6.69 and HGX H100 $3.99; 1-Click Clusters B200 $9.86 and H100 $6.16 per GPU-hour at 16 GPUs.
