VALDI
VALDI is an LA-based GPU cloud platform offering on-demand and reserved access to NVIDIA and AMD accelerators (H100, A100, L40S, GH200, MI250) across 11 data centers in 8 countries. It serves AI/ML researchers, universities, and enterprise AI teams with pay-as-you-go compute, detachable storage, and developer APIs.
- Company typePrivate
- Founded2022
- HeadquartersLos Angeles, United States
- Headcount11–50
- GTM typeB2B
- OfferingSoftware
What VALDI does
VALDI operates a global GPU cloud infrastructure platform headquartered in Pasadena, California, delivering on-demand and reserved access to high-performance NVIDIA and AMD accelerators (H100, A100, L40S, RTX A6000, GH200, H200, MI250) for AI training and inference workloads. Its infrastructure spans 11 data centers across 8 countries (USA, Norway, UAE, India, Czech Republic, Canada, Australia), anchored by US DC-1 with 2,048 NVIDIA H100 80GB SXM5 cards and up to 600MW of raw power capacity. The platform is built on kernel-level virtualization via KVM and QEMU for VM isolation, with public REST API access, a Python SDK, and one-click deployment scripts for popular AI models including Llama 3.x and Stable Diffusion. The company was acquired by Storj in July 2024 and operates as VALDI LABS, PBC, a Delaware-incorporated Public Benefit Corporation.
The business model combines a product-led growth motion (self-serve sign-up with a $10 minimum deposit, per-second billing, no contracts or minimums) with sales-led reserved instances and an enterprise post-pay option. Pricing is positioned at roughly 50% below hyperscalers for GPU compute and 57% below AWS S3 for storage, with a 10% educational discount applied automatically for .edu sign-ups. Revenue streams include usage-based GPU compute (on-demand and reserved), cloud storage at $0.01/GB-month for detachable volumes, enterprise post-pay arrangements, educational discounts, and a $5 referral credit incentive. Distribution is augmented by a third-party reseller program with single-tenant API architecture and a GPU List Marketplace where verified suppliers can list inventory.
Customer base is anchored by named academic institutions including Caltech, UIUC, USC, Northeastern, Harvard, and Yale, alongside enterprise AI teams and research labs running workloads on frontier models. The platform supports AI/ML engineers and researchers with on-demand training and inference, scalable detachable storage with AES-256 encryption and erasure coding, training burst clusters for multi-GPU jobs, and dedicated developer tooling (CLI, API, GitHub scripts). Co-founder and CEO Nikhil Jain leads the company, which maintains compliance certifications including ISO 27001, SOC 1 Type II, SOC 2 Type II, and PCI DSS across select data center facilities.
VALDI firmographics
Firmographics- Name
- VALDI
- Legal name
- VALDI LABS, PBC
- Website
- https://valdi.ai
- Company type
- Private
- Founded year
- 2022
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- VALDI is an LA-based GPU cloud platform offering on-demand and reserved access to NVIDIA and AMD accelerators (H100, A100, L40S, GH200, MI250) across 11 data centers in 8 countries. It serves AI/ML researchers, universities, and enterprise AI teams with pay-as-you-go compute, detachable storage, and developer APIs.
- Ownership category
- akta.pro rank
VALDI industry classification
Industry- Product category
- GPU Cloud Infrastructure
- NAICS
- Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (51821), Computer Facilities Management Services (541513)
- SIC
- Services-Computer Programming, Data Processing, Etc. (7370), Services-Computer Processing & Data Preparation (7374)
- akta.pro primary industry
- AI Compute Cloud & GPU-as-a-Service (HDAAAAAK)
- akta.pro secondary industry
- GPU-Accelerated & AI Training/Inference Servers (HDACABAG)
Keywords
Where VALDI is headquartered
LocationHeadquarters
- HQ city
- Los Angeles
- HQ country
- United States
- HQ region
- North America
Offices1 record
Markets served
VALDI business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Infrastructure, Technology or R&D, Personnel, Operations, Marketing or Sales
Revenue model
- GPU Compute (On-Demand): Pay-as-you-go model where customers are billed by the second for GPU compute usage. Customers deposit funds and spin up compute resources; no contracts or minimums. Starting at $0.22/hr for RTX A4000 up to $2.35/hr for NVIDIA H100.
- GPU Compute (Reserved/Long-term): Customers can reserve GPU instances for longer terms (month-to-month, 3 months, 6 months, 12 months, 24 months) at discounted rates. Requires upfront payment ranging from 1 week to 12 weeks depending on commitment length.
- Cloud Storage: Scalable, redundant network storage at $0.11/GB-month for additional storage, with $0.01/GB-month for standard detachable volumes and $0.01/GB for data downloads. 57% cheaper than AWS S3.
- Enterprise Post-Pay: Post-pay arrangements offered to enterprise and university customers, enabling employees or students to use VALDI without pre-paying with a credit card. Requires direct contact with support team.
- Educational Discounts: Automatic 10% discount applied to all services when signing up with a .edu email address, with prices automatically reflecting the discount.
- Referral Program: Customers earn $5 in credits when a referred user deposits at least $10, driving customer acquisition through existing user referrals.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Usage-based | Pay-as-you-go | On-demand GPU access with instant availability and per-second billing |
| Usage-based | Multi-year contract | Reserved GPU instances at discounted rates with upfront payment commitments |
| Usage-based | Pay-as-you-go | Global cloud storage at competitive rates |
| Subscription | Monthly | 10% discount for verified educational users |
Go-to-market motion4 records
Distribution channels4 records
Marketing channels7 records
VALDI product offering
Product offeringCore offering
VALDI operates a GPU cloud computing platform that provides on-demand and reserved GPU instances for AI and machine learning workloads, alongside companion cloud object storage. The service is designed to deliver high-performance NVIDIA GPU compute at a claimed lower cost than traditional hyperscalers, with global data center reach, instant provisioning, and flexible reservation terms for AI researchers, universities, and enterprise teams.
Product overview
VALDI is an LA-based cloud GPU compute platform offering on-demand and reserved access to NVIDIA and AMD GPUs for AI workloads. The core offering consists of On-Demand GPU Compute and GPU Reservation services, supplemented by VALDI Storage (detachable encrypted volumes), VALDI CLI and API for programmatic access, and specialized products like Llama-as-a-Service and Stable Diffusion deployment guides. The platform supports AI training and inference through various GPU types (H100, A100, L40S, GH200, RTX A6000, H200) across global data centers in the US, Australia, Canada, Czech Republic, India, Norway, and UAE. Additional services include Training Bursts for intensive workloads, Enterprise Post-Pay options, a Reseller Program, and a GPU List Marketplace for verified suppliers.
Differentiator
Problem solved
Functional benefit
Products and services
- On-Demand GPU Compute
Quantifiable outcome
- 50% lower GPU cost vs hyperscalers
- +4 more outcomes
Companies that use VALDI
Customer profileNamed customers7 records
Segments4 records
Ideal customer profiles3 records
VALDI technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
AI capability5 records
Feature5 records
VALDI partnerships and signals
Strategic signalScale indicators5 records
Recent moves6 records
Expansion highlights6 records
VALDI competitors and assessment
Company assessmentDirect peers
- CoreWeave: CoreWeave is a specialized GPU cloud provider offering on-demand and reserved NVIDIA H100/A100 instances for AI training and inference. It is the most direct competitor to VALDI, operating at significantly larger scale with similar positioning around flexible GPU compute for AI workloads.
- Lambda Labs: Lambda operates GPU cloud clusters built on NVIDIA hardware (H100, A100, GH200) targeted at AI researchers and enterprises, alongside its own Lambda Inference API. Overlaps with VALDI on pay-as-you-go and reserved GPU offerings for AI training and inference.
- Vast.ai: Vast.ai runs a marketplace that matches AI/ML workloads to idle consumer and prosumer GPUs globally, mirroring VALDI's marketplace model for matching ML jobs to spare compute. Both target price-sensitive AI researchers with flexible, lower-cost GPU access.
- RunPod: RunPod offers on-demand and reserved GPU cloud pods (H100, A100, RTX 4090) for AI training and inference, with both instant serverless endpoints and dedicated instances. Competes directly with VALDI on per-second billing, low minimums, and developer-first PLG motion.
- Together AI: Together AI provides GPU cloud infrastructure for training and serving open-source LLMs, with dedicated inference APIs. Comparable to VALDI in serving AI researchers and developers with on-demand GPU access plus higher-level model-serving products.
- Fluidstack: Fluidstack is a GPU cloud provider offering large-scale AI compute clusters (H100, H200, GH200) on flexible terms to AI labs and enterprises. Comparable to VALDI in the reserved-instance AI compute space, though typically at larger enterprise contract sizes.
- Paperspace (DigitalOcean): Paperspace, now part of DigitalOcean, provides on-demand GPU cloud instances and notebooks for AI/ML developers. Comparable to VALDI on developer-focused GPU instances, though priced closer to hyperscaler rates and bundled with DigitalOcean's broader cloud portfolio.
Broad incumbents
- Amazon Web Services (EC2 P5 instances): AWS offers P5/P4d EC2 instances powered by H100/A100 GPUs as part of its broader cloud portfolio. It is the incumbent competitor VALDI benchmarks its 50% lower pricing against, but lacks VALDI's specialization and pay-as-you-go simplicity for individual AI developers.
- Google Cloud (A3 instances): Google Cloud provides A3 instances with H100 GPUs alongside Vertex AI and broader data services. Competes with VALDI for enterprise AI workloads but bundles GPUs with Vertex AI's managed ML platform rather than offering raw, developer-friendly GPU access.
Emerging players
- Crusoe Energy: Crusoe builds AI cloud infrastructure powered by stranded/renewable energy, offering H100 GPU clusters to AI labs. Comparable to VALDI in providing cost-advantaged GPU compute, differentiated by energy source and target customer profile of frontier AI labs.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat4 records
Key risks6 records
Key highlights7 records
Customer concentration
VALDI social profiles
Digital presenceVALDI compliance and trust
Trust signalCompliance10 records
VALDI financial estimates
Financial estimateRevenue estimate
Valuation estimate
VALDI leadership team
Management profileNumber of profiles
Profiles1 record
VALDI funding detail
Funding detailFunding overview
Funding rounds
Investors
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
VALDI M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about VALDI
What does VALDI do?
VALDI operates a GPU cloud computing platform that provides on-demand and reserved GPU instances for AI and machine learning workloads, alongside companion cloud object storage. The service is designed to deliver high-performance NVIDIA GPU compute at a claimed lower cost than traditional hyperscalers, with global data center reach, instant provisioning, and flexible reservation terms for AI researchers, universities, and enterprise teams.
Is VALDI a public or private company?
VALDI is a private company. It is classified as corporate owned and is currently operating.
When was VALDI founded?
VALDI was founded in 2022. It employs 11 to 50 people.
Where is VALDI based?
VALDI is headquartered in Los Angeles, United States, in the North America region.
How does VALDI make money?
Six revenue lines are on record. GPU Compute (On-Demand) is the primary driver. The others are GPU Compute (Reserved/Long-term), cloud Storage, enterprise Post-Pay, educational Discounts and referral Program.
Who are VALDI's main competitors?
Direct peers on record are CoreWeave, Lambda Labs, Vast.ai, RunPod, Together AI, Fluidstack and Paperspace (DigitalOcean). Broad incumbents are Amazon Web Services (EC2 P5 instances) and Google Cloud (A3 instances). Crusoe Energy is listed as an emerging player.
Does VALDI have an API?
Yes. VALDI provides a public API for full integration capabilities. The API supports reseller arrangements with single-tenant architecture where third-party resellers manage users under a primary account. Authentication uses JWT tokens that expire every 5 minutes. API key support is planned but not yet available. Developer documentation is at docs.valdi.ai/api.
What industry is VALDI in?
VALDI's product category is GPU Cloud Infrastructure. Its primary akta.pro industry code is HDAAAAAK, AI Compute Cloud & GPU-as-a-Service, with a secondary code of HDACABAG, GPU-Accelerated & AI Training/Inference Servers. Its NAICS code is 518 and its SIC code is 7370.