
Shadeform
6/10Shadeform is a GPU cloud marketplace. AI teams use one console, API and CLI to compare, launch and manage GPU instances across 20 to 30-plus cloud providers.
Data last updated November 2026 · Our methodology
Overview
Shadeform is a marketplace and control plane for renting GPUs. It does not own any hardware. Instead, it gathers on-demand and reserved GPU capacity from many neoclouds, including Lambda, Nebius, Crusoe, Voltage Park, DigitalOcean, Vultr and Verda. Customers find, launch and manage instances from one console, REST API, shade CLI or MCP server, and billing and support are handled centrally. Instances can attach persistent volumes and can run a startup script or container at launch. Access is over SSH. The vendor says capacity spans more than 100 regions in North America, Europe, Asia and Australia, and that all GPUs sit in Tier 3 or Tier 4 data centers. On pricing, Shadeform's FAQ says it charges the same hourly rate the underlying provider charges and adds no fee for launching instances or managing accounts. Standard billing is pay-as-you-go, and invoicing is available on request. Reserved commitments get provider quotes through a request form. Shadeform also brokers large bare-metal deals and sells managed clusters under the name Shadeform Capacity Solutions. It suits AI startups, inference providers and ML teams that want to compare GPU prices and availability without opening accounts at many clouds. Its main limitation is that it depends on the underlying providers. SemiAnalysis's ClusterMAX 2.0 review rated it Bronze and noted that it has no monitoring dashboards, cannot run health checks on provider machines, and had no testable Slurm or Kubernetes cluster product. Machine SLAs also come from each provider, not from Shadeform.
Our Rating
AI Software6.1/10
Hosting & Infrastructure5.9/10
Who Shadeform is best for
Best software for StartupsGood fit
Its early customers were mainly AI-native startups, and it lets them buy GPUs pay-as-you-go without long-term commitments.
It offers an API, a CLI, an MCP server and SSH access for ML engineers.
Inference providers and AI product companies use it to source GPU capacity across several clouds.
Less suited
It reports Fortune 500 and hyperscaler customers, but its monitoring and health checks are limited.
Pros & Cons
Pros
- Its FAQ says it charges the provider's own rate with no platform markup.
- One account and one bill cover GPUs from many neoclouds.
- SemiAnalysis found it had the widest provider coverage of any GPU marketplace it reviewed, with 23 providers at the time of testing.
- Developers get an API, the shade CLI and an MCP server, and instances can run startup scripts or containers at launch.
- Listings show compliance details such as SOC 2 Type II and HIPAA for each provider.
Cons
- SemiAnalysis noted that it has no monitoring dashboards.
- It cannot run health checks on the underlying provider machines.
- Uptime SLAs come from each provider, so Shadeform offers no uniform SLA of its own.
- It is a small company with limited funding, reported at about 2 million dollars.
- Support can be inconsistent, according to the ClusterMAX review.
Key Features
Shadeform Pricing
On-demand (pay-as-you-go)
- No platform fee
- Centralized billing
- GPU and CPU instances from many providers
- Same hourly rate as the underlying provider
Reserved Commitments
- Reservation discounts for longer terms
- Provider quotes at no cost or obligation
- Choose GPU model, interconnect, quantity, term and location
Frequently Asked Questions
No. Its FAQ says it charges the same rate as going directly to the cloud provider, with no extra fee.
Yes. Billing is pay-as-you-go by default, and you can email support to arrange invoice payment.
Machine SLAs are inherited from the underlying cloud providers, and specific SLAs can be shared on request.
Email support@shadeform.ai, which the team says it monitors closely, or join its Slack community.
Yes. Submit a Reserved Commitments request and Shadeform returns provider quotes at no cost or obligation.
Over SSH as the shadeform user, using the key the instance was launched with. Instances can be managed through the console, REST API, CLI or MCP server.
Similar Tools
See all alternatives
Bare metal cloud provider that offers dedicated servers, GPU servers and virtual machines on demand, billed hourly or monthly, for developers and infrastructure-heavy businesses.

Luxembourg-based edge and cloud provider offering CDN, virtual machines, bare metal, managed Kubernetes, object storage, AI GPU cloud, and DDoS/WAF security for businesses.

CoreWeave is a GPU-focused cloud that rents NVIDIA GPU and CPU instances, managed Kubernetes, and storage to teams training and serving AI models.
Ranked near Shadeform in AI Software
See Shadeform in the listShadeform is #505 in AI Software by our score.





