Spheron ABU
Spheron ABU operates a GPU cloud marketplace aggregating enterprise-grade NVIDIA capacity from certified Tier 3/4 data centers worldwide, offering on-demand, spot, and reserved rentals to AI startups, research labs, and enterprises at 40-60% below hyperscaler pricing.
- Company typePrivate
- Founded2020
- HeadquartersTel Aviv, Israel
- Headcount1–10
- GTM typeB2B
- OfferingSoftware
What Spheron ABU does
Spheron ABU is a Tel Aviv-based GPU cloud marketplace that aggregates enterprise-grade NVIDIA capacity from certified Tier 3/4 data centers worldwide into a single platform. The company traces its operating history to a 2020 founding with $7M of disclosed initial funding, secured additional 2022 strategic rounds led by Alpha Wave Global with participation from ConsenSys Mesh, CoinDCX Ventures, Zee Prime Capital, Tykhe Ventures, iSeed, 3one4 Capital, and Paradigm Shift Capital, and operates under the Spheron Network brand at spheron.network. Inventory spans 13+ NVIDIA models from RTX 4090 ($0.53/hr) and A100 through H100, H200, B200, B300, plus rack-scale GB200 and GB300 NVL72 systems with 1,024+ GPUs available and pre-orders for the Rubin R100 shipping H2 2026. Partner data centers hold ISO 27001, SOC 2 Type I/II, and HIPAA certifications across North America, Europe, and Asia Pacific.
Core products comprise on-demand dedicated instances (99.99% SLA), spot instances at up to 50% discount using idle partner capacity, and reserved commitments running one week to 12+ months for 8-512+ GPU clusters with InfiniBand interconnect and dedicated setup support. Per-minute billing starts and stops on instance state changes with no minimum run time, and the self-serve platform at app.spheron.ai provisions bare-metal SSH access with Docker and CUDA pre-installed in under two minutes. Complementary proprietary tooling includes a GPU Recommender that maps any Hugging Face model to cheapest compatible hardware and a Training Cost Calculator; programmatic deployment and billing are exposed via a documented REST API and an open-source GitHub presence (spheron-core). Mitrasish Mukherjee is identified as co-founder and CTO.
Revenue is generated through usage-based on-demand rentals, margin on spot utilization, and recurring reserved-capacity enterprise contracts, with payment options spanning enterprise invoicing and USDC/USDT stablecoins. Customers include AI/ML startups (Prem AI, Baseten, Eigen Labs, Exo Labs, Cognichip, NeevCloud, Hosted.ai, Wafer, Ionet, Fluence, Faraway, Humanify, Artic Digital), academic institutions (Stanford, UC Berkeley, University of Washington), and consumer AI enterprises (Replika). Go-to-market combines a product-led self-serve motion targeting AI developers with an enterprise field-sales team handling 100+ GPU cluster procurements.
Spheron ABU firmographics
Firmographics- Name
- Spheron ABU
- Legal name
- Spheron Network
- Website
- https://spheron.network
- Company type
- Private
- Founded year
- 2020
- Operating status
- Operating
- Headcount range
- 1–10 employees
- Short description
- Spheron ABU operates a GPU cloud marketplace aggregating enterprise-grade NVIDIA capacity from certified Tier 3/4 data centers worldwide, offering on-demand, spot, and reserved rentals to AI startups, research labs, and enterprises at 40-60% below hyperscaler pricing.
- Ownership category
- akta.pro rank
Spheron ABU industry classification
Industry- Product category
- Cloud GPU Infrastructure
- NAICS
- Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182)
- SIC
- Services-Computer Programming, Data Processing, Etc. (7370)
- akta.pro primary industry
- AI Compute Cloud & GPU-as-a-Service (HDAAAAAK)
- akta.pro secondary industries
- GPU/Accelerator Compute Servers (HDABADAC), GPU-Accelerated & AI Training/Inference Servers (HDACABAG)
Keywords
Where Spheron ABU is headquartered
LocationHeadquarters
- HQ city
- Tel Aviv
- HQ country
- Israel
- HQ region
- Middle East
Offices1 record
Markets served
Spheron ABU business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Infrastructure, Technology or R&D, Personnel, Marketing or Sales, Operations
Revenue model
- GPU Rental (On-Demand): Per-minute billing for dedicated GPU instances with 99.99% SLA. Customers pay only for minutes used with no minimum commitment or warm-up charges. Rates vary by GPU model (A100 from $0.82/hr, H100 from $2.01/hr, B200 from $5.34/hr).
- GPU Rental (Spot Instances): Discounted GPU rental using idle capacity from data center network. Same hardware as dedicated instances but interruptible when demand rises. Typical savings of 30-60% off dedicated rates.
- Reserved GPU Commitments: Long-term GPU reservations for weeks to 12+ months with guaranteed capacity and volume pricing. Designed for large deployments (8-512+ GPUs) with custom cluster configurations and dedicated support.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Unit Pricing | Pay-as-you-go | RTX 4090 - Budget development and experimentation |
| Unit Pricing | Pay-as-you-go | A100 - General ML workloads |
| Unit Pricing | Pay-as-you-go | H100 - LLM training and large-scale inference |
| Unit Pricing | Pay-as-you-go | H200 - Long-context LLM inference |
| Unit Pricing | Pay-as-you-go | B200 - Large-model training and FP4 inference |
| Unit Pricing | Pay-as-you-go | B300 - Frontier training and trillion-parameter models |
| Subscription | Multi-year contract | Reserved Capacity - Custom clusters |
Go-to-market motion3 records
Distribution channels3 records
Marketing channels8 records
Spheron ABU product offering
Product offeringCore offering
Spheron operates a GPU cloud marketplace that aggregates enterprise-grade GPU capacity from certified Tier 3/4 data centers worldwide into a single, transparent platform. Customers rent NVIDIA H100, H200, B200, B300, A100, and RTX PRO 6000 instances with per-minute billing starting at $0.53/hr, deploying in under 2 minutes without contracts. The service includes spot instances at up to 50% discount, reserved capacity for 8-512+ GPU clusters, and a unified API for programmatic instance management.
Product overview
Spheron is a GPU cloud marketplace that aggregates enterprise-grade GPU capacity from certified Tier 3/4 data centers worldwide into a single, transparent platform. The core offering is on-demand GPU rental (H100, H200, B200, B300, A100, RTX PRO 6000 from $0.53-$5.34/hr) with per-minute billing and no contracts, supplemented by spot instances (up to 50% discount) and reserved capacity commitments (1 week to 12+ months) for enterprise customers. Rack-scale NVL72 systems (GB200, GB300) provide 72-GPU clusters with unified memory for trillion-parameter training. Complementary tools include the GPU Recommender (finds cheapest GPU for HuggingFace models) and Training Cost Calculator (estimates training cost/time). All instances provide bare-metal access with SSH root, Docker, and CUDA pre-installed.
Differentiator
Problem solved
Functional benefit
Products and services
- On-Demand GPU Rental
Quantifiable outcome
- 40-60% cost savings vs hyperscalers (H100 annual: Spheron $14,400 vs AWS $39,456)
- +4 more outcomes
Companies that use Spheron ABU
Customer profileNamed customers18 records
Segments4 records
Ideal customer profiles4 records
Spheron ABU technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
AI capability11 records
Feature4 records
Spheron ABU partnerships and signals
Strategic signalPartnerships
One partnership is on record.
- Certified Data Center Partners (Tier 3/4)coreSpheron aggregates GPU capacity from certified Tier 3/4 data centers worldwide. Partners provide infrastructure with ISO 27001, SOC 2 Type I & II, and HIPAA certifications. All facilities have redundant power, cooling, and network infrastructure.
Scale indicators6 records
Recent moves6 records
Expansion highlights5 records
Spheron ABU competitors and assessment
Company assessmentEmerging players
- Crusoe: Vertically integrated AI cloud provider building owned GPU clusters in energy-advantaged locations, recently expanding into managed AI inference. Represents a competing model to Spheron's asset-light aggregation, with deeper infrastructure ownership as a potential long-term threat.
- Akash Network: Decentralized compute marketplace for GPU and CPU workloads, token-incentivized and globally distributed. Pursues the same multi-provider aggregation thesis as Spheron in the decentralized/crypto-native segment, increasingly serving AI training customers.
Direct peers
- Together AI: AI cloud platform combining GPU infrastructure (H100/H200 clusters) with managed inference and fine-tuning services. Overlaps with Spheron in GPU rental for LLM workloads while adding a higher-level platform layer that Spheron could eventually need to match.
- Lambda Labs: GPU cloud and on-demand cluster provider (H100, H200, GH200) targeting AI startups and research labs with self-serve and reserved offerings. Overlaps with Spheron in self-serve GPU rental, cost positioning, and AI-native customer base.
- RunPod: GPU cloud platform offering on-demand and serverless GPU pods (H100, A100) targeted at AI developers and startups with per-second billing. Spheron and RunPod compete head-to-head for the same price-sensitive AI developer and startup segment with similar self-serve motions.
- Vast.ai: Marketplace for peer-to-peer and data center GPU rentals offering per-hour and per-month pricing on consumer and enterprise GPUs. Operates a similar aggregation model with transparent pricing and no long-term contracts, directly overlapping with Spheron's on-demand value proposition.
- CoreWeave: GPU cloud provider offering H100, H200, and Blackwell instances with reserved capacity for AI training and inference. Spheron explicitly benchmarks its H100 pricing against CoreWeave ($14,400 vs $18,000 annual), positioning the two as direct cost-alternative competitors for AI infrastructure buyers.
- io.net: Decentralized GPU network aggregating distributed compute capacity for AI workloads. io.net is even listed as a Spheron customer/partnership contact, and pursues the same marketplace aggregation model Spheron uses to source capacity from third-party data centers.
- FluidStack: GPU cloud provider operating large-scale AI compute clusters (H100, H200, Blackwell) with reserved capacity and managed offerings. Competes with Spheron in the reserved-cluster and enterprise segment where Spheron offers 8-512+ GPU configurations.
Broad incumbents
- Paperspace (DigitalOcean): GPU cloud offering (now part of DigitalOcean) providing on-demand and reserved H100/A100 instances targeting developers and startups. Competes with Spheron for the AI developer segment but with the backing of a larger parent (DigitalOcean) and broader cloud portfolio.
Market position
Strengths4 records
Weaknesses4 records
Competitive moat4 records
Key risks5 records
Key highlights6 records
Customer concentration
Spheron ABU social profiles
Digital presenceSpheron ABU compliance and trust
Trust signalCompliance4 records
Spheron ABU financial estimates
Financial estimateRevenue estimate
Valuation estimate
Spheron ABU leadership team
Management profileNumber of profiles
Profiles1 record
Spheron ABU funding detail
Funding detailFunding overview
Funding rounds3 records
Investors13 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Spheron ABU 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 Spheron ABU
What does Spheron ABU do?
Spheron operates a GPU cloud marketplace that aggregates enterprise-grade GPU capacity from certified Tier 3/4 data centers worldwide into a single, transparent platform. Customers rent NVIDIA H100, H200, B200, B300, A100, and RTX PRO 6000 instances with per-minute billing starting at $0.53/hr, deploying in under 2 minutes without contracts. The service includes spot instances at up to 50% discount, reserved capacity for 8-512+ GPU clusters, and a unified API for programmatic instance management.
Is Spheron ABU a public or private company?
Spheron ABU is a private company. It is classified as venture growth investor backed and is currently operating.
When was Spheron ABU founded?
Spheron ABU was founded in 2020. It employs 1 to 10 people.
Where is Spheron ABU based?
Spheron ABU is headquartered in Tel Aviv, Israel, in the Middle East region.
How does Spheron ABU make money?
Three revenue lines are on record. GPU Rental (On-Demand) is the primary driver. The others are GPU Rental (Spot Instances) and reserved GPU Commitments.
Who are Spheron ABU's main competitors?
Emerging players on record are Crusoe and Akash Network. Direct peers are Together AI, Lambda Labs, RunPod, Vast.ai, CoreWeave, io.net and FluidStack. Paperspace (DigitalOcean) is listed as a broad incumbent.
Does Spheron ABU have an API?
Yes. Spheron offers API access through their developer documentation at docs.spheron.ai. The API reference allows developers to programmatically interact with the platform for GPU instance management, deployment, and billing operations. The platform supports per-minute billing APIs for cost tracking and management. Developer documentation is at docs.spheron.ai/api-reference.
What industry is Spheron ABU in?
Spheron ABU's product category is Cloud GPU Infrastructure. Its primary akta.pro industry code is HDAAAAAK, AI Compute Cloud & GPU-as-a-Service, with a secondary code of HDABADAC, GPU/Accelerator Compute Servers. Its NAICS code is 5182 and its SIC code is 7370.