What Hosted·AI does

Hosted·AI is a software company founded in 2024 and headquartered in San Jose, California, that builds a turnkey GPU-as-a-Service (GPUaaS) orchestration platform for service providers. The company targets neocloud operators, cloud service providers, telecommunications companies, and hosting providers — entities that own GPU infrastructure but lack turnkey software to monetize it as elastic, multi-tenant GPU services. Its core technology is a software-defined GPU architecture that pools physical GPUs into virtual pools and enables configurable overcommit ratios (2x-10x), adaptive temporal and spatial multi-tenant scheduling, and unified provisioning across GPU services, virtual machines, and bare metal instances. The platform claims up to 5x improvement in GPU utilization versus the neocloud industry average of approximately 40%, and supports both NVIDIA accelerators and FuriosaAI's RNGD Tensor Contraction Processor.

The company monetizes through a usage-based software licensing model priced per gigabyte-hour of VRAM managed ($0.0005/GB-h) and consumed ($0.0030/GB-h), with a $750 monthly minimum commitment and no upfront licensing fees. Professional services for go-to-market advisory, deployment assistance, and financing/hardware procurement facilitation (including commission dynamics via Maerifa Solutions) supplement the core software revenue. The go-to-market combines a self-service product-led motion (publicly accessible ROI calculators, GPU Picker, GPU Price Index, and Business Planner tools) with sales-assisted enterprise engagement via demo booking, layered with a channel partnership for neocloud creation and owned demand-generation properties — packet.ai (a consumer-facing neocloud sub-brand) and GPUaaS.com (a wholesale GPU matchmaking marketplace). Public launch occurred in January 2025; the company closed a $19M seed round in March 2026 led by Creandum, bringing total funding to approximately $23.63M.

Name

Hosted·AI

Legal name

Hosted AI Inc

Website

https://hosted.ai

Company type

Private

Founded year

2024

Operating status

Operating

Headcount range

11–50 employees

Short description

Hosted·AI is a San Jose-based software company that provides a turnkey GPU-as-a-Service orchestration platform for service providers, including neoclouds, cloud service providers, telcos, and hosting providers. Its software-defined GPU architecture enables multi-tenant pooling, overcommit, and adaptive scheduling across NVIDIA and Furiosa accelerators.

Ownership category

akta.pro rank

Product category

GPU Cloud Infrastructure Software

NAICS

Software Publishers (51321), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518)

SIC

Services-Prepackaged Software (7372)

akta.pro primary industry

AI Compute Virtualization & Scheduling (GPU virtualization, cluster schedulers) (HDAAAAAG)

akta.pro secondary industry

AI Compute Cloud & GPU-as-a-Service (HDAAAAAK)

Keywords

• GPU orchestration software
• Multi-tenant GPUaaS
• AI infrastructure platform
• Neocloud software
• GPU resource pooling

Headquarters

HQ city

San Jose

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

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

Revenue model

1. GPUaaS Platform Licensing: Software licensing fees based on VRAM managed by the platform (availability fee of $0.0005/GB-h) and VRAM consumed by customers ($0.0030/GB-h). Providers pay based on the GPU VRAM under management, not per GPU card. No upfront licensing fees. 24×7 support and 15-minute SLA included.
2. Professional Services & GTM Advisory: The company offers go-to-market advisory, deployment assistance, and support to help service providers launch and optimize their GPUaaS offerings.
3. Financing & Hardware Procurement Facilitation: Through financial partners and Maerifa, hosted.ai facilitates financing options and hardware procurement for service providers building neoclouds, creating a commission/referral revenue dynamic.

Pricing tiers

Go-to-market motion2 records

Distribution channels5 records

Marketing channels7 records

Core offering

Hosted·AI sells a turnkey GPU-as-a-Service (GPUaaS) software platform that enables service providers to pool physical GPUs into virtual GPU pools, provision multi-tenant AI workloads, and oversubscribe GPU resources by 2x–10x for higher utilization and margin. The platform is hardware-agnostic (supporting NVIDIA and Furiosa RNGD accelerators), ships with billing integration (WHMCS), a brandable self-service neocloud portal, and a full REST API, and is licensed on a usage-based model tied to VRAM managed and consumed.

Product overview

hosted·ai is a turnkey software platform for GPU-as-a-Service (GPUaaS) designed specifically for service providers including neoclouds, CSPs, telcos, and hosts. The portfolio consists of the core hosted·ai platform (GPUaaS software with GPU pooling, multi-tenant scheduling, and overcommit), supplemented by packet·ai (a neocloud sub-brand offering GPUaaS at competitive prices), GPUaaS.com (a wholesale GPU matchmaking service), and a suite of neocloud business planning tools (Business Planner, Funding + Cashflow Planner, Marketing ROI, GPU Price Index, GPU Picker). The platform also includes GPU Mesh for buying and selling spare GPU capacity without additional CAPEX. The architecture enables up to 5x GPU utilization improvement, configurable overcommit ratios (2x-10x), and supports NVIDIA and Furiosa accelerators.

Differentiator

Problem solved

Functional benefit

Brands

• packet.ai: A neocloud powered by hosted.ai customers' optimized GPU infrastructure, delivering GPUaaS at market-leading prices.
• GPUaaS.com

Products and services

• hosted·ai GPUaaS Software Platform Turnkey GPUaaS software platform that enables service providers to pool physical GPUs into virtual pools, provision multi-tenant AI workloads, and oversubscribe GPU resources by 2x–10x for up to 5x utilization improvement. Includes software-defined GPU control, billing and metering, WHMCS integration, brandable self-service customer portal, REST API, and unified provisioning across GPU services, VMs, and bare metal. Targets neocloud operators, CSPs, telcos, and hosting providers.
• packet·ai A neocloud sub-brand of Hosted·AI that delivers GPUaaS at market-leading prices by leveraging GPU infrastructure from hosted·ai platform customers. Generates direct demand for hosted·ai-powered neoclouds and serves as an open-source reference implementation for the neocloud portal.
• GPUaaS.com A wholesale GPU matchmaking service (sub-brand of Hosted·AI) connecting enterprise GPU buyers with hosted·ai platform customers and partners that can fulfill custom GPU cluster requirements at scale. Routes demand to hosted·AI-powered neoclouds.

Quantifiable outcome

• Up to 5x GPU utilization improvement, reducing CAPEX requirements 5x
• +4 more outcomes

Named customers6 records

Segments3 records

Ideal customer profiles2 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration1 record

Feature9 records

Partnerships

Two partnerships are on record, tiered core.

• Maerifa SolutionscoreChannel Partner/ Reseller/ Distributor · 30 September 2025Strategic partnership to provide a one-stop shop for neocloud creation at scale. Maerifa provides technology design, deployment, and supply chain management services, with relationships across NVIDIA, Supermicro, Lenovo, and finance partners. Combined with hosted.ai, this enables rapid neocloud creation with turnkey software for orchestration and monetization. Maerifa is ADGM-registered and operates across Europe, Middle East, Africa, and Southeast Asia. Already working on multiple joint projects at time of announcement.
• FuriosaAIcoreTechnology or Integration · 8 July 2025Strategic partnership to deliver AI infrastructure powered by Furiosa's Tensor Contraction Processor (TCP) architecture. The hosted.ai platform will support Furiosa's flagship RNGD processor for LLM and agentic AI inference. Furiosa's TCP chip architecture is purpose-built for AI inference workloads, delivering high performance per watt. The partnership enables service providers to leverage TCP-powered infrastructure with full multi-tenant virtualization and oversubscription via hosted.ai software. Expected availability by end of 2025, with plans for an off-the-shelf appliance combining hosted.ai software, Furiosa accelerators, and rack server modules.

Scale indicators4 records

Recent moves6 records

Expansion highlights7 records

Direct peers

• Vast.ai: Decentralized GPU compute marketplace connecting GPU owners with AI developers. Both target AI compute buyers with competitive pricing; Vast.ai aggregates third-party GPU supply while hosted.ai enables professional providers to operate GPUaaS offerings.
• Crusoe Cloud: GPU cloud provider utilizing stranded energy for AI compute infrastructure. Both serve AI developers needing GPU compute; Crusoe operates its own neocloud while hosted.ai provides the software platform that powers comparable neoclouds.
• Paperspace: GPU cloud computing platform (acquired by DigitalOcean) offering virtual machines and notebooks for AI/ML workloads. Both provide GPUaaS to developers and ML practitioners; Paperspace is owned by a broader cloud provider pursuing similar AI compute positioning.
• Lambda: GPU cloud and AI infrastructure provider offering GPU instances for training and inference. Comparable as both sell GPU compute services to AI developers and enterprises, though Lambda operates its own fleet while hosted.ai enables other providers to do so.
• RunPod: GPU cloud platform offering on-demand GPU pods and serverless AI compute. Both serve AI developers needing GPU resources with self-service portals; RunPod is a competing GPUaaS provider using its own orchestration layer.
• CoreWeave: Large-scale GPU cloud provider offering GPU compute services for AI workloads. Both CoreWeave and hosted.ai target the AI compute market; CoreWeave operates the cloud itself while hosted.ai provides the orchestration software used by similar providers.

Emerging players

• Together AI: AI cloud platform offering GPU-backed infrastructure for training and inference, with focus on open-source models. Both serve AI compute buyers with optimized GPU stacks; Together AI is vertically integrated with model serving while hosted.ai is horizontal infrastructure software.
• Render: Cloud platform offering GPU instances for AI/ML and rendering workloads with developer-friendly pricing. Both serve GPU compute buyers; Render focuses on developer experience while hosted.ai's software enables similar experiences across multiple providers.

Broad incumbents

• Vultr: Global cloud infrastructure provider offering cloud compute, GPUs, bare metal, and hosting. Vultr is a competitor in the GPU cloud space but operates as a broad cloud platform rather than a dedicated GPUaaS provider, mirroring the CSP/Telco/Host customer segment hosted.ai targets.
• OVHcloud: European cloud provider offering GPU instances alongside broader cloud services. Comparable as a CSP expanding into GPUaaS, which is the exact customer segment hosted.ai targets for its orchestration software.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks5 records

Key highlights7 records

Customer concentration

• LinkedInlinkedin.com/company/hostedai

Revenue estimate

Valuation estimate

Number of profiles

Profiles4 records

Funding overview

Funding rounds1 record

Investors7 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 Hosted·AI