Crusoe AI Data Center

What Crusoe AI Data Center does

Crusoe AI Data Center is a vertically integrated AI infrastructure company that designs, builds, and operates energy-secured AI factories for hyperscale and AI-native customers. Founded in 2021 (originally as Crusoe Energy Systems, pivoting from flare gas-powered bitcoin mining to AI infrastructure), the company is headquartered in Denver, Colorado (with offices in San Francisco, Dublin, Tel Aviv, and other locations) and operates under the legal entity Crusoe Technologies LLC. The platform combines proprietary energy generation and procurement (renewables, stranded gas, second-life EV batteries via Redwood Materials, advanced nuclear via Blue Energy, iron-air batteries via Form Energy) with modular and hyperscale data center development (the 1.2 GW Abilene flagship, plus 4.9 GW of total contracted capacity across five campuses and a 40+ GW development pipeline) and a vertically integrated AI cloud platform (Crusoe Cloud) offering NVIDIA GB200/B200/H200/H100 and AMD MI300X/MI355X GPUs, Managed Inference powered by proprietary MemoryAlloy technology, Crusoe Managed Kubernetes, Crusoe Edge Zones, and Crusoe Spark modular data centers manufactured at the Brighton, Colorado Spark Factory.

Crusoe's business model is built on two complementary revenue streams: long-term, take-or-pay hyperscale infrastructure contracts (Microsoft 900 MW in Abilene, Meta 1.6 GW across Childress and Warrenton, Oracle and OpenAI on the Stargate Abilene campus) which anchor utilization and provide revenue predictability, and a usage-based Crusoe Cloud platform (GPU-as-a-Service and Managed Inference) serving AI-first startups and enterprise developers with self-serve console access and API-first delivery. Pricing is not publicly disclosed, but customer-reported evidence indicates 50% cost savings versus traditional hyperscalers and 99.98% cluster uptime. The company has raised over $4 billion in cumulative capital (most recently a $1.375B Series E at a $10B+ valuation co-led by NVIDIA in October 2025, with a $3B round reportedly in discussion at a $30B valuation as of July 2026), and operates a $15 billion joint venture with Blue Owl Capital and Primary Digital Infrastructure for data center development.

Crusoe's core differentiation is its energy-first vertical integration: from patented flare gas mitigation (Digital Flare Mitigation) to on-site power generation, in-house modular manufacturing, and proprietary inference optimization software. This stack targets the binding constraint of AI infrastructure — power availability and speed-to-deployment — and is supported by partnerships with NVIDIA (strategic investor and technology partner), Lancium (Abilene Clean Campus), Tallgrass (Wyoming), Energy Vault (Texas), Baker Hughes and Bergen Engines (generator supply), and channel partners including atNorth (Nordic), Anan (Israel), and Saturn Cloud (enterprise AI development). Major customers include Oracle, OpenAI, Meta, Microsoft, Windsurf, Cursor, Figure, Cartesia, SES AI, and a broad PLG base of AI startups, with deployments across the United States, Iceland, Ireland, Israel, and the Philippines.

Name

Crusoe AI Data Center

Legal name

Crusoe Technologies LLC

Website

https://crusoe.ai

Company type

Private

Founded year

2021

Headcount range

1,001–5,000 employees

Short description

Crusoe AI Data Center is a vertically integrated AI infrastructure company that builds, powers, and operates energy-secured AI factories and GPU cloud services (Crusoe Cloud, Managed Inference, Crusoe Spark modular units) for hyperscale anchors like Microsoft, Meta, Oracle, and OpenAI and AI-native startups globally.

Ownership category

akta.pro rank

Product category

Cloud AI Infrastructure

NAICS

Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518), Computer Facilities Management Services (541513)

SIC

Services-Prepackaged Software (7372), Services-Computer Integrated Systems Design (7373), Services-Computer Programming, Data Processing, Etc. (7370)

akta.pro primary industry

AI Datacenter Power, Cooling & Racks (liquid cooling, power delivery) (HDAAAAAH)

akta.pro secondary industries

End-to-End Enterprise AI Platforms (MLOps & Model Lifecycle Management) (HDAEANAA), AI Observability, Monitoring & Evaluation Platforms (Drift, Quality, Safety) (HDAEANAF)

Keywords

• AI cloud infrastructure
• GPU compute services
• Modular data centers
• Managed AI inference
• AI factory operations

Headquarters

HQ city

Denver

HQ country

United States

HQ region

North America

Offices15 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Infrastructure, Technology or R&D, Personnel, Supply Chain, Operations, Marketing or Sales

Revenue model

1. AI Cloud Infrastructure (GPU as a Service): Provides GPU compute resources (NVIDIA GB200, B200, H200, H100, AMD MI300X, MI355X) on-demand through Crusoe Cloud platform. Customers pay for compute usage with peak performance features including accelerated storage and optimized RDMA networking.
2. Managed Inference Services: Fully managed AI inference platform offering access to top open-source models (Llama, DeepSeek, Nemotron, Kimi, Qwen, Gemma) and custom fine-tuned models. Proprietary MemoryAlloy technology delivers 9.9x faster inference performance.
3. AI Data Center Development: Design, construction, and operation of modular and hyperscale AI data centers for major enterprise clients including Microsoft, Oracle, OpenAI, and Meta. Revenue from long-term infrastructure contracts and colocation agreements.
4. Modular AI Factory Sales/Leasing: Sale and deployment of Crusoe Spark modular AI data centers (~1MW units) to enterprise customers. Includes rapid deployment within three months and flexible configuration options.

Pricing tiers

Go-to-market motion3 records

Distribution channels4 records

Marketing channels8 records

Core offering

Crusoe AI Data Center operates a vertically integrated AI infrastructure platform that designs, builds, and operates large-scale AI data centers and provides Crusoe Cloud, a GPU-as-a-service platform with on-demand access to NVIDIA and AMD accelerators. The company pairs modular data center manufacturing (Crusoe Spark) with cloud services including Managed Inference, Command Center, and Edge Zones, delivering energy-first AI factories at gigawatt scale for hyperscalers and AI developers.

Product overview

Crusoe AI Data Center operates as a vertically integrated AI infrastructure company offering a platform-plus-modules architecture. The core product is Crusoe Cloud, an AI cloud platform providing infrastructure as a service with high-performance NVIDIA (GB200, B200, H200, H100) and AMD (MI300X, MI355X) GPUs. Key modules include Managed Inference (powered by proprietary MemoryAlloy technology for 9.9x faster inference), Crusoe Intelligence Foundry (for deploying open-source models), Command Center (unified operations), and Crusoe Edge Zones/Spark (modular edge AI data centers). Supporting infrastructure includes Managed Kubernetes, Managed Slurm, and AutoClusters. The company also manufactures modular AI data centers at its Spark Factory facility.

Differentiator

Problem solved

Functional benefit

Brands

• Crusoe Cloud: AI cloud platform providing GPU infrastructure-as-a-service, including managed Kubernetes, managed Slurm, and fault-tolerant AutoClusters.
• Crusoe Spark
• Crusoe Edge Zones
• Crusoe Managed Inference
• Crusoe Intelligence Foundry
• Crusoe Command Center

Products and services

• Crusoe Cloud AI cloud platform providing GPU infrastructure-as-a-service with high-performance NVIDIA and AMD compute, accelerated storage, and optimized RDMA networking designed for AI developers, AI startups, and enterprises running training and inference workloads.
• Managed Inference Managed AI inference service powered by proprietary MemoryAlloy technology, delivering up to 9.9x faster time-to-first-token for production AI workloads across open-source and custom fine-tuned models.
• Crusoe Edge Zones Modular AI data centers delivering low-latency, sovereign, and rapidly deployable AI infrastructure at virtually any location, leveraging MemoryAlloy technology for accelerated inference performance.
• Crusoe Spark Modular AI data center units of approximately one megawatt each, manufactured in-house and sold or leased to enterprise customers for rapid scalable edge AI deployments deployable within three months.
• Spark Factory Manufacturing facility in Brighton, Colorado producing Crusoe Spark modular AI factory units at scale, with capacity of approximately 100 modular units per year.
• Fastokens Open-source Rust-based BPE tokenizer optimized for AI inference workloads, built in collaboration with NVIDIA Dynamo and integrated into SGLang by LMSYS, delivering up to 50% improvement in time-to-first-token for agentic AI workloads.
• AI Data Center Development Services End-to-end design, construction, and operation of modular and hyperscale AI data center campuses for major enterprise clients under multi-gigawatt long-term infrastructure contracts.

Quantifiable outcome

• 9.9x faster time-to-first-token for AI inference workloads
• +4 more outcomes

Named customers9 records

Segments3 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration3 records

AI capability6 records

Feature8 records

Partnerships

20 partnerships are on record, tiered core, major and minor.

• NvidiacoreTechnology or Integration · 4 July 2026Major technology and strategic investor partner; Nvidia provides GPU technology (GB200, B200, H200, H100) and has invested in Crusoe. Nvidia CEO Jensen Huang involved in Crusoe's AI factory initiatives. Crusoe is early adopter of NVIDIA's latest AI infrastructure technologies including Vera CPU and Nemotron models.
• Meta PlatformscoreStrategic or Co-development Partner · 19 June 2026Meta secured 1.6 GW AI computing capacity agreements at Childress, Texas and Warrenton, Missouri facilities. Equivalent to powering 750,000 US homes.
• Bergen EnginesmajorTechnology or Integration · 3 June 2026~750 MW U.S. power agreement for AI data centers. Gas-powered generator solutions for reliable power generation at Crusoe facilities.
• AnanmajorStrategic or Co-development Partner · 27 April 2026$1 billion deal to establish and operate 40-100 MW server farm at Anan's data center in Afula, Israel. First step in broader collaboration to build global AI infrastructure platform from Israel.
• MicrosoftcoreStrategic or Co-development Partner · 27 March 2026Microsoft secured 900 MW AI factory campus development in Abilene, Texas. First building expected to open mid-2027. Microsoft's Azure cloud integration with Crusoe infrastructure.
• Form EnergymajorStrategic or Co-development Partner · 24 March 202612 gigawatt-hours of iron-air batteries for AI data centers. Batteries can store and discharge electricity for up to 100 hours, enabling long-duration energy storage without relying on constrained grids.
• Blue EnergymajorStrategic or Co-development Partner · 14 March 2026Commitment to build hyperscale data centers in Clark, Pampanga, Philippines as part of broader AI infrastructure expansion.
• Baker HughesmajorTechnology or Integration · 1 March 202625 BRUSH Power Generation electric generators (1.21 GW total) for Crusoe AI data centers from mid-2026 through 2028. Deal underscores Baker Hughes' push into power solutions for AI infrastructure.
• atNorthmajorChannel Partner/ Reseller/ Distributor · 24 February 2026Expansion of Iceland data center capacity with atNorth partnership. Nordic colocation services and heat reuse partnerships.
• Energy VaultmajorStrategic or Co-development Partner · 11 February 2026Framework agreement for phased deployment of Crusoe Spark modular AI data centers at Energy Vault's Texas technology center. Initial program scalable up to 25 MW.
• Blue EnergymajorStrategic or Co-development Partner · 30 October 2025Partnership to develop advanced nuclear-powered AI data center project in Port of Victoria, Texas. Integrating nuclear and renewable energy sources.
• StarcloudminorStrategic or Co-development Partner · 22 October 2025Strategic partnership to become first cloud operator in space, expanding into satellite-based cloud computing.
• TallgrassmajorStrategic or Co-development Partner · 24 July 2025Joint venture for AI data center in Wyoming with dedicated gas and renewable power sources. Tallgrass provides energy infrastructure development.
• GE VernovamajorTechnology or Integration · 22 July 202529-unit aeroderivative gas turbine deal to deliver power to AI data centers. Major power infrastructure partnership.
• Redwood MaterialscoreStrategic or Co-development Partner · 26 June 2025Expansion of partnership to scale renewable-powered AI compute. Deployment scaled from 4 to 24 Crusoe Spark modular data centers at Redwood's Nevada campus. Second-life EV battery microgrid (12 MW / 63 MWh) with solar achieving 99.2% availability.
• Engine No. 1majorStrategic or Co-development Partner · 17 March 2025Powered data center infrastructure joint venture combining Engine No. 1's investment strategy with Crusoe's infrastructure development.
• OpenAIcoreStrategic or Co-development Partner · 1 January 2025Primary anchor tenant for Crusoe Abilene Stargate campus. Part of $500 billion Stargate AI project. OpenAI utilizing hundreds of thousands of GPUs on single integrated network fabric.
• LanciumcoreStrategic or Co-development Partner · 18 July 2024Real estate and energy technology partner for Abilene Clean Campus development. Lancium provides land acquisition, grid interconnect, site engineering, renewables interconnect, and power orchestration. 1.2 GW grid interconnection at Abilene site.
• OraclecoreTechnology or Integration · 1 June 2024Oracle Cloud Infrastructure running on Crusoe Abilene data center campus. Oracle delivers NVIDIA GB200 racks to the facility. Part of broader 4.5 GW Stargate AI infrastructure agreement for OpenAI.
• RescaleminorChannel Partner/ Reseller/ Distributor · 16 January 2024Partnership for sustainable accelerated computing on Crusoe Cloud through Rescale's AI-powered R&D platform.

Scale indicators14 records

Recent moves13 records

Expansion highlights8 records

Direct peers

• CoreWeave: Largest pure-play NVIDIA GPU cloud provider offering H100/B200 clusters and managed inference to AI labs and enterprises — directly comparable to Crusoe Cloud in target customer, GPU SKU mix, and large-context inference workload focus.
• Lambda: GPU cloud provider delivering NVIDIA clusters to AI startups, research labs and enterprises with on-demand and reserved capacity — competes head-to-head with Crusoe Cloud for AI-first startups and offers overlapping managed GPU services.
• Applied Digital: Designs, builds and operates gigawatt-scale AI data center campuses with long-term anchor tenant leases — closest peer to Crusoe's JV-based hyperscale data center development model and renewable power siting strategy.
• Nebius Group: Vertically integrated AI infrastructure operator combining proprietary GPU cloud with purpose-built data centers in Europe and the US, targeting AI startups and frontier-model customers — directly comparable business model and product breadth.

Broad incumbents

• Vantage Data Centers: Global hyperscale and enterprise colocation provider with multi-gigawatt campuses across North America and EMEA — overlaps with Crusoe's hyperscale build-and-lease motion but is broader/colocation-focused rather than AI-native.
• Digital Realty: Largest global data center REIT operating 300+ facilities across 50+ metros — comparable as a major AI-ready colocation incumbent, but focused on general-purpose wholesale and retail colocation rather than vertically integrated AI factories.
• Equinix: Global interconnection-focused colocation operator with dense cloud on-ramps — comparable in scale and global data center footprint, but serves a broader enterprise interconnection/colo mix rather than Crusoe's AI-factory positioning.
• Iron Mountain Data Centers: Hyperscale-ready colocation provider repurposing legacy infrastructure into AI-suitable capacity — competes for the same anchor tenants and large footprint leases, but is a broader incumbent rather than an AI-native operator.

Emerging players

• TeraWulf: Crypto-mining operator pivoting to AI/HPC hosting with owned behind-the-meter power and contracted hyperscale customers — shares Crusoe's energy-first heritage and Bitcoin-to-AI transition arc, but remains smaller and earlier-stage.

Others

• Penguin Solutions: Provides accelerated computing infrastructure, cluster design and managed services for AI and HPC workloads (including NVIDIA-based systems) — comparable to the systems-integration and managed-cluster layer of Crusoe's stack rather than competing on bare GPU rental.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

• Xx.com/CrusoeAI
• GitHubgithub.com/crusoecloud
• LinkedInlinkedin.com/company/crusoe-ai

Compliance2 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles15 records

Subsidiaries1 record

Funding overview

Funding rounds3 records

Investors8 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Crusoe AI Data Center