Lambda

Lambda, founded in 2012 by brothers Stephen and Michael Balaban, operates what it calls the "Superintelligence Cloud"—a GPU infrastructure platform for AI model training and inference.

Reviewed by 7wData

On this page

Profile

Operates GPU cloud infrastructure and data centers where companies rent on-demand or reserved NVIDIA GPUs to train and serve AI models.

Lambda, founded in 2012 by brothers Stephen and Michael Balaban, operates what it calls the "Superintelligence Cloud"—a GPU infrastructure platform for AI model training and inference. The company evolved from early work in facial recognition APIs to become a major player in the competitive GPU cloud market, competing directly with CoreWeave (which IPO'd in 2025). Lambda built its reputation on developer experience: pre-configured instances with TensorFlow, PyTorch, and other ML frameworks, avoiding the DevOps complexity competitors impose.

Today, Lambda Stack software runs on 50,000+ ML teams globally, and the platform serves over 5,000 direct cloud customers ranging from researchers to Fortune 500 enterprises. The funding arc accelerated sharply in 2024–2026. A $320 million Series C (February 2024, led by Thomas Tull's USIT) valued the company at $1.5 billion.

By February 2025, a $480 million Series D came from Andra Capital and SGW at roughly $2.5 billion. Then, in November 2025, Lambda announced a multibillion-dollar infrastructure contract with Microsoft and simultaneously closed a $1.5 billion Series E led by TWG Global (Tull and Mark Walter), reaching a $5.9–6 billion valuation. In January 2026, the company raised another $350 million in pre-IPO convertible notes led by Mubadala Capital, with an H2 2026 IPO targeted.

Beyond Microsoft, NVIDIA itself became a major customer: Lambda signed a $1.5 billion lease-back deal to absorb 18,000 of NVIDIA's own chips. At NVIDIA GTC 2026 (March), Lambda announced Bare Metal Instances and became a launch partner for NVIDIA's Vera CPU and STX deployments. The company counts Microsoft, Google, Amazon, Intel, and Tencent among its enterprise users, plus MIT, Stanford, Harvard, and the Department of Defense.

Leadership remains intact: Stephen Balaban as CEO, Michael Balaban now in a Chief Product Officer role, and John Donovan (former AT&T Communications CEO) as board chair. Glassdoor reviews (3.4/5 stars, 31 reviews) paint a mixed culture: engineering rigor and a sharp team, but reports of tech-bro dynamics and insufficient response to harassment allegations.

Track Lambda and 240+ vendors.

335k+ subscribers read the daily AI & data note. One email, both newsletters. Unsubscribe anytime.

Who buys this

  • Large enterprises training foundation models (Microsoft, Google, Amazon, Tencent)
  • AI research institutions and universities (MIT, Stanford, Harvard, Caltech, Los Alamos)
  • AI startups and independent ML researchers and developers
  • Government agencies and defense contractors
  • Organizations running inference and LLM serving at scale

Publicly disclosed clients

  • Microsoft
  • Google
  • Amazon
  • Intel
  • Tencent
  • MIT
  • Stanford
  • Harvard
  • Caltech
  • Los Alamos National Lab
  • Disney
  • Department of Defense
  • Kaiser Permanente

Strengths and what to watch

Strengths

  • Pre-configured ML platform with TensorFlow, PyTorch, and frameworks out-of-the-box, reducing DevOps friction compared to competitors.
  • Secured multibillion-dollar contracts with Microsoft (Nov 2025) and a $1.5B lease-back from NVIDIA, anchoring revenue and capital availability.
  • Strong capital position ($3.1B+ raised, $5.9–6B valuation) and institutional backing from Tull, Walter, T. Rowe Price, and Mubadala Capital positions Lambda for scale.

Watch for

  • IPO execution in H2 2026; market receptivity to GPU cloud infrastructure plays depends on broader sentiment around AI capex and Nvidia competition.
  • Glassdoor culture rating (3.4/5 stars) with employee reports of harassment, lack of psychological safety, and diversity concerns suggest retention and hiring risks.
  • Customer concentration: NVIDIA is largest single customer ($1.5B lease-back); Microsoft multibillion-dollar deal suggests heavy dependency on hyperscaler demand and cloud strategy shifts.

Recent moves

Key Information

Industry
Machine Learning & Artificial Intelligence
Founded
2012
Headquarters
United States

Frequently Asked Questions

What is Lambda?

Lambda operates the Superintelligence Cloud, a GPU infrastructure platform founded in 2012. The company rents on-demand and reserved NVIDIA GPUs to companies training and serving AI models. Lambda Stack software runs on 50,000+ ML teams globally, serving 5,000+ direct customers from startups to Fortune 500 enterprises.

How does Lambda help train AI models?

Lambda provides pre-configured GPU instances with TensorFlow, PyTorch, and other ML frameworks built-in, eliminating DevOps complexity. Companies choose on-demand or reserved GPU capacity for model training and inference at scale. The platform handles infrastructure management, letting teams focus on AI development rather than server configuration.

Which companies use Lambda?

Lambda serves Microsoft, Google, Amazon, Intel, and Tencent among enterprises, plus MIT, Stanford, Harvard, Caltech, and the Department of Defense. The company signed a multibillion-dollar infrastructure agreement with Microsoft in November 2025 and a $1.5 billion lease-back deal with NVIDIA, anchoring its revenue and capital position.

How much funding has Lambda raised?

Lambda has raised over $3.1 billion total across 13 funding rounds. Its Series E in November 2025 brought in $1.5 billion, valuing the company at $5.9–6 billion. Additionally, the company raised $350 million in pre-IPO convertible notes in January 2026, with an H2 2026 IPO currently targeted.

What makes Lambda different from other GPU cloud providers?

Lambda's key advantage is pre-configured ML instances with TensorFlow, PyTorch, and frameworks built-in, reducing DevOps friction. Unlike competitors, the platform eliminates complex infrastructure setup. Lambda Stack software running on 50,000+ teams globally demonstrates its developer-first approach, though it competes directly with CoreWeave in the growing GPU cloud market.

Is Lambda a good place to work?

Glassdoor reviews rate Lambda 3.4 out of 5 stars based on 31 reviews. Employees report strong engineering quality but note tech-bro dynamics, insufficient response to harassment allegations, and diversity concerns. The culture feedback suggests retention and hiring risks, though the company maintains strong backing and growth trajectory.

How Lambda compares

Direct head-to-head against 3 competitors. Picked by 7wData.

This company

Lambda

Positioning
Operates GPU cloud infrastructure and data centers where companies rent on-demand or reserved NVIDIA GPUs to train and serve AI models.
Customer segments
Large enterprises training foundation models (Microsoft, Google, Amazon, Tencent)
Strengths
Pre-configured ML platform with TensorFlow, PyTorch, and frameworks out-of-the-box, reducing DevOps friction compared to competitors.
Watch for
IPO execution in H2 2026; market receptivity to GPU cloud infrastructure plays depends on broader sentiment around AI capex and Nvidia competition.
Recent moves
Lambda announces multibillion-dollar agreement with Microsoft to deploy AI infrastructure powered by tens of thousands of NVIDIA GPUs

CoreWeave

Positioning
Dedicated GPU cloud for large AI labs and enterprises; IPO'd Nasdaq March 2025; 43 data centers, 250,000-plus owned GPUs.
Customer segments
AI labs (OpenAI, Meta), enterprises needing reserved hyperscale GPU capacity; buyers on multi-year committed contracts.
Strengths
First cloud to deploy NVIDIA GB200 NVL72 instances at scale; $22.4B in committed contracts with OpenAI.
Watch for
Microsoft represented over 60% of 2024 revenue; customer concentration risk detailed in IPO filings.
Recent moves
Closed $8.5B investment-grade GPU-backed financing facility, April 2026; first investment-grade rating for GPU-secured debt.

Crusoe

Positioning
Energy-first GPU cloud company; builds large-scale AI data centers anchored to low-cost, low-carbon power sources.
Customer segments
Enterprise AI teams and startups requiring large reserved GPU clusters; buyers where energy cost and sustainability matter.
Strengths
$15B joint venture with Blue Owl for a 1.2-GW Abilene, TX AI campus; largest energy-anchored GPU buildout in the US.
Watch for
Texas grid crisis in March 2025 spiked energy costs 40%; power price volatility is a structural margin risk.
Recent moves
$1.38B Series E at $10B valuation, October 2025, led by Founders Fund and NVIDIA.

Lightning AI

Positioning
Merged GPU cloud and ML development platform; $2.5B-plus combined valuation; 35,000-plus owned NVIDIA H100, B200, and GB300 GPUs.
Customer segments
Developers, ML researchers, and enterprises seeking on-demand GPU access; 400,000-plus users across training and production workloads.
Strengths
35,000-plus owned NVIDIA GPUs at publicly listed on-demand rates; PyTorch Lightning framework used by 400,000-plus active developers.
Watch for
Voltage Park's primary backer was Ripple (XRP creator); crypto-linked capital provenance draws scrutiny from enterprise procurement teams.
Recent moves
Completed merger of Lightning AI and Voltage Park, January 21 2026, forming combined entity valued at over $2.5B.

Sources

  1. techcrunch.com — Series E $1.5B funding details, Microsoft deal context, valuation trajectory
  2. www.businesswire.com — NVIDIA Vera CPU partnership, STX deployment, Bare Metal Instances launch, Q1 2026 product announcements
  3. www.crunchbase.com — Total funding ($3.1B+), number of funding rounds (13), founding year (2012), key investors (USIT, TWG Global, Andra, SGW, T. Rowe Price)
  4. www.glassdoor.com — Glassdoor rating (3.4/5 stars), culture feedback (tech-bro dynamics, harassment concerns, diversity issues), employee sentiment (58% would recommend)
  5. lambda.ai — Product offerings (Superclusters, 1-Click Clusters, Bare Metal Instances), mission statement, customer segments, Lambda Stack software
  6. techcrunch.com — Microsoft multibillion-dollar agreement (November 2025), NVIDIA GB300 NVL72 deployment, strategic partnership details