Vast.ai
Vast.ai operates a two-sided GPU cloud marketplace connecting 1,300+ hardware providers with 120,000+ developers and AI teams, offering API-native access to 20,000+ GPUs across 40+ data centers at 60-80% below hyperscaler pricing via self-serve, startup, and enterprise tiers.
- Company typePrivate
- Founded2016
- HeadquartersLos Angeles, United States
- Headcount11–50
- GTM typeB2B
- OfferingDigital Commerce or Content
What Vast.ai does
Vast.ai operates a two-sided GPU cloud marketplace that connects 1,300+ hardware providers with 120,000+ developers and AI teams needing compute. The platform aggregates 20,000+ GPUs across 40+ data centers and 68+ GPU types, including consumer cards (RTX 4090, RTX 5090) and enterprise silicon (H100, H200, B200, AMD GPUs). Prices are set by real-time supply and demand rather than vendor markup, producing a 60-80% cost advantage versus AWS, GCP, and Azure. The platform is API-native: provisioning, lifecycle management, billing, and serverless orchestration are exposed via REST API, Python SDK, and CLI, enabling AI agents to programmatically design, procure, and optimize compute. Per-second billing with a $5 minimum and no long-term contracts characterize the self-serve PLG motion, complemented by an enterprise field-sales tier for bulk pricing, SOC 2 Type II compliant isolated clusters, and HIPAA/GDPR-friendly configurations, plus a startup credit-matching program.
Three core deployment products anchor the offering: GPU Cloud for on-demand instances, Serverless for auto-scaling inference endpoints with predictive optimization, and Clusters for dedicated multi-node training with InfiniBand networking. Supporting products include a Model Library of pre-configured open-source model templates, a Hosting module that lets GPU owners monetize idle hardware, Vast Finance for GPU equipment financing, and Vast Hardware for vetted procurement. Revenue is generated primarily through marketplace commissions on each rental transaction, with subscription-style enterprise contracts adding a secondary recurring stream. The company was founded in 2016 in Los Angeles by Jake Cannell and Christian Horne, is incorporated as a Delaware C Corporation, and currently operates with 40+ employees across LA and San Francisco offices, having achieved SOC 2 Type I certification in April 2025 and 310% year-over-year revenue growth in 2024.
Vast.ai firmographics
Firmographics- Name
- Vast.ai
- Legal name
- Vast.ai Inc.
- Website
- https://vast.ai
- Company type
- Private
- Founded year
- 2016
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- Vast.ai operates a two-sided GPU cloud marketplace connecting 1,300+ hardware providers with 120,000+ developers and AI teams, offering API-native access to 20,000+ GPUs across 40+ data centers at 60-80% below hyperscaler pricing via self-serve, startup, and enterprise tiers.
- Ownership category
- akta.pro rank
Vast.ai industry classification
Industry- Product category
- GPU Cloud Marketplace
- NAICS
- Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518)
- SIC
- Services-Computer Rental & Leasing (7377), Services-Computer Processing & Data Preparation (7374)
- akta.pro primary industry
- AI Compute Cloud & GPU-as-a-Service (HDAAAAAK)
- akta.pro secondary industries
- Decentralized GPU/AI Compute Networks (training/inference, GPU renting, model hosting) (FSAPALAC), Model Hosting, Serving & Inference Platforms (HDAAACAB), AI Compute Virtualization & Scheduling (GPU virtualization, cluster schedulers) (HDAAAAAG)
Keywords
Where Vast.ai is headquartered
LocationHeadquarters
- HQ city
- Los Angeles
- HQ country
- United States
- HQ region
- North America
Offices2 records
Markets served
Vast.ai business model
Business model- GTM type
- B2B
- Offering type
- Digital Commerce or Content
- Cost components
- Technology or R&D, Personnel, Infrastructure, Operations, Marketing or Sales
Revenue model
- GPU Rental Marketplace: Vast.ai operates a two-sided marketplace connecting GPU renters (developers, AI teams) with GPU hosts (data centers, individual owners). Takes a commission on each rental transaction. Hosts set their own prices and Vast handles customer acquisition, support, and billing.
- GPU Hosting Platform Services: Revenue from hosting services where Vast handles customer support, billing, and platform operations while hosts provide hardware. Platform fee applied to transactions.
- Enterprise/Bulk GPU Contracts: Custom pricing agreements for enterprises with large-scale, long-term workloads including volume discounts and reserved GPU contracts.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Usage-based | Pay-as-you-go | On-Demand GPU instances with guaranteed uptime |
| Usage-based | Pay-as-you-go | Preemptible instances at 50%+ discount |
| Usage-based | Multi-year contract | Reserved capacity with up to 50% savings |
| Subscription | Multi-year contract | Enterprise bulk pricing for large-scale deployments |
| Hybrid | Monthly | Startup credit matching program |
Go-to-market motion3 records
Distribution channels4 records
Marketing channels7 records
Vast.ai product offering
Product offeringCore offering
Vast.ai operates a two-sided GPU cloud marketplace connecting 1,300+ GPU providers with renters via an API-native platform. It offers on-demand GPU instances (GPU Cloud), serverless inference with autoscaling, and dedicated multi-node Clusters across 40+ data centers and 68+ GPU types. The platform charges per-second billing, sets prices by supply and demand, and supports programmatic provisioning via CLI, Python SDK, and REST API for AI/ML training, inference, fine-tuning, and agentic workloads.
Product overview
Vast.ai is a GPU cloud marketplace platform offering three primary deployment products: GPU Cloud for on-demand instances, Serverless for auto-scaling inference endpoints, and Clusters for dedicated multi-node training. The platform connects GPU providers (hosts) with 120K+ developers needing compute, supporting 68+ GPU types across 40+ data centers. Supporting products include a Model Library of pre-configured templates, Hosting for GPU owners to earn revenue, Enterprise solutions with compliance and bulk pricing, Startup Program with credit matching, CLI/Python SDK developer tools, and Vast Finance/Hardware for GPU procurement and financing.
Differentiator
Problem solved
Functional benefit
Products and services
- GPU Cloud On-demand GPU instance rental service providing access to 20,000+ GPUs across 40+ data centers worldwide with per-second billing and support for 68+ GPU types including RTX 4090, H100, H200, and B200. Deploy in seconds via CLI, SDK, or API.
- Serverless Serverless GPU inference platform enabling deployment of models as endpoints with automatic benchmarking and optimization across GPU types. Autoscales to zero with predictive optimization for cost and latency; pay only for compute time.
- Clusters Dedicated multi-node GPU clusters with InfiniBand networking for large-scale AI training workloads, offering scalable performance, enterprise SLAs, and 24/7 support.
- Model Library Pre-configured templates and ready-to-deploy models including Kimi K2.6, Qwen3.6 35B, Gemma 4 31B, FLUX.2, and LTX-2.3 for text generation, image generation, video generation, audio generation, and computer vision.
- Hosting GPU hosting platform enabling users to list and rent out their GPU hardware. Hosts set their own prices, choose rental types (on-demand, interruptible, reserved), and earn revenue from 120K+ developers on the platform.
- Enterprise Enterprise-grade GPU infrastructure solutions including bulk pricing, compliance-ready isolated clusters, private networking, VPN access, white-glove support, and custom software configurations for large-scale deployments.
- Startup Program Credit matching program that matches startup compute spend dollar-for-dollar up to set amounts, enabling companies to double their GPU compute budget during growth phases.
- Vast Finance GPU hardware financing service connecting hosts with financing partners who specialize in GPU infrastructure, offering equipment-as-collateral leases, revenue-based lending, and traditional equipment financing.
- Vast Hardware GPU procurement service sourcing new and certified refurbished GPU servers through a vetted supplier network. Hardware is burn-in tested, benchmark-verified, and pre-configured for the Vast platform.
Quantifiable outcome
- 60% GPU cost reduction while scaling to 200K daily users
- +3 more outcomes
Companies that use Vast.ai
Customer profileNamed customers10 records
Segments5 records
Ideal customer profiles4 records
Vast.ai technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
AI capability12 records
Feature7 records
Vast.ai partnerships and signals
Strategic signalPartnerships
Two partnerships are on record, tiered flagship and core.
- NGCGflagshipNGCG announced strategic expansion into AI infrastructure leveraging partnerships with NVIDIA, Dell Technologies, Newegg, Best Buy, and Vast.ai to deploy scalable AI compute for recurring revenue through hourly, short-term, and contract-based leasing models.
- NVIDIAcoreVast.ai is listed as an ecosystem partner for NVIDIA DSX Air, a SaaS platform for digital simulation of AI factory infrastructure. Vast demonstrates integrations for validating full-stack AI infrastructure environments before hardware deployment.
Scale indicators10 records
Recent moves8 records
Expansion highlights6 records
Vast.ai competitors and assessment
Company assessmentDirect peers
- CoreWeave: CoreWeave is the most directly comparable GPU cloud provider, offering on-demand NVIDIA GPU compute (H100, H200, B200) for AI training and inference. It competes head-to-head with Vast.ai's GPU Cloud and Clusters products, though CoreWeave owns/operates its data centers rather than running a marketplace.
- Lambda Labs: Lambda provides GPU cloud instances and dedicated clusters for AI workloads with similar target customers (AI researchers, startups). Closely comparable on product offering (on-demand H100s, clusters) and developer-focused GTM, though Lambda is owned-operator rather than marketplace.
- RunPod: RunPod offers GPU cloud rentals and serverless AI endpoints targeting developers and AI startups, with a marketplace-like aggregation across multiple data centers. Direct competition on pricing, GPU variety, and per-second billing.
- Fluidstack: Fluidstack provides large-scale GPU compute for AI training and inference, including dedicated clusters and hyperscale contracts. Competes with Vast's enterprise/cluster offering, particularly for foundation model labs.
- io.net: io.net is a decentralized GPU compute network (DePIN) that aggregates underutilized GPUs for AI training and inference. Closely comparable to Vast.ai's marketplace model and agent-ready positioning, with a crypto-incentivized supply side.
Emerging players
- Together AI: Together AI offers GPU cloud for training/inference with emphasis on open-source models. Overlaps with Vast's Model Library and serverless products; competes for the same AI startup workload segment.
- Crusoe Cloud: Crusoe provides GPU cloud compute using stranded/renewable energy, competing in the AI training and inference space with a sustainability angle. Comparable to Vast on GPU rental and cost-efficient positioning.
Broad incumbents
- Paperspace (DigitalOcean): Paperspace (acquired by DigitalOcean) offers GPU cloud instances for ML workloads, now part of a broader cloud platform. Overlaps with Vast on developer-focused GPU rental but DigitalOcean offers much wider portfolio.
- Amazon Web Services (EC2/EKS/ECS GPU): AWS is the dominant cloud provider with P4/P5 GPU instances for AI workloads. While not a direct marketplace peer, it is the primary incumbent that Vast's cost arbitrage positions against.
Regional players
- Genesis Cloud: Genesis Cloud offers GPU cloud instances with a European focus and sustainability angle. Comparable on developer-focused GPU rental and cost-efficient pricing, but with a different geographic emphasis.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat6 records
Key risks6 records
Key highlights7 records
Customer concentration
Vast.ai social profiles
Digital presenceVast.ai compliance and trust
Trust signalCompliance4 records
Vast.ai financial estimates
Financial estimateRevenue estimate
Valuation estimate
Vast.ai leadership team
Management profileNumber of profiles
Profiles6 records
Vast.ai funding detail
Funding detailFunding overview
Funding rounds1 record
Investors1 record
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Vast.ai 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 Vast.ai
What does Vast.ai do?
Vast.ai operates a two-sided GPU cloud marketplace connecting 1,300+ GPU providers with renters via an API-native platform. It offers on-demand GPU instances (GPU Cloud), serverless inference with autoscaling, and dedicated multi-node Clusters across 40+ data centers and 68+ GPU types. The platform charges per-second billing, sets prices by supply and demand, and supports programmatic provisioning via CLI, Python SDK, and REST API for AI/ML training, inference, fine-tuning, and agentic workloads.
Is Vast.ai a public or private company?
Vast.ai is a private company. It is classified as venture growth investor backed and is currently operating.
When was Vast.ai founded?
Vast.ai was founded in 2016. It employs 11 to 50 people.
Where is Vast.ai based?
Vast.ai is headquartered in Los Angeles, United States, in the North America region.
How does Vast.ai make money?
Three revenue lines are on record. GPU Rental Marketplace is the primary driver. The others are GPU Hosting Platform Services and enterprise/Bulk GPU Contracts.
Who are Vast.ai's main competitors?
Direct peers on record are CoreWeave, Lambda Labs, RunPod, Fluidstack and io.net. Emerging players are Together AI and Crusoe Cloud. Broad incumbents are Paperspace (DigitalOcean) and Amazon Web Services (EC2/EKS/ECS GPU). Genesis Cloud is listed as a regional player.
Does Vast.ai have an API?
Yes. REST API providing programmatic access to search offers, create instances, manage resources, track billing, and build custom automation. Uses bearer token authentication via Authorization header or query parameter. Available at https://cloud.vast.ai/api/v0/ with OpenAPI spec available. Developer documentation is at docs.vast.ai/api-reference/introduction.
What industry is Vast.ai in?
Vast.ai's product category is GPU Cloud Marketplace. Its primary akta.pro industry code is HDAAAAAK, AI Compute Cloud & GPU-as-a-Service, with a secondary code of FSAPALAC, Decentralized GPU/AI Compute Networks (training/inference, GPU renting, model hosting). Its NAICS code is 5182 and its SIC code is 7377.