Thunder Compute logo

GPU Cloud Provider · Unknown

Thunder Compute

Thunder Compute provides cost-effective, on-demand GPU instances tailored towards AI/ML workloads. It offers a developer-friendly environment with rapid deployment and flexible hardware configurations, emphasizing a smooth user experience through features like persistent storage and VS Code integration.

Visit Provider View 3 GPUs
GPUs
3
Founded
Unknown

GPU Marketplace

NVIDIA RTX A6000On-Demand
$0.35/hour
SpecsDeploy
NVIDIA L40On-Demand
$0.79/hour
SpecsDeploy
NVIDIA H100 PCIeOn-Demand
$3.20/hour
SpecsDeploy

Estimates only — rates are collected automatically from public provider pages and may be out of date. Prices vary by region, commitment term, and availability, and typically exclude storage, egress, and tax. Confirm current pricing with the provider before purchasing. Last collected .

Company Profile

Company TypeA GPU cloud platform for developers.
Provider TypeA GPU cloud platform for developers.
Legal EntityThunder Compute

Infrastructure

GPU FleetRTX A6000, L40, A100 80GB, H100 PCIe
Network FabricDynamic IP, 7 Gbps Egress/Ingress
Connectivity7-10 Gbps
StoragePersistent storage with snapshots
Data Center TierEnterprise-grade data centers
AvailabilityAvailable
Data scientists, startups, and researchers

Compute & Deployment

On-DemandBilled per minute
VM-BasedUses proprietary virtualization software

GPU Hardware

Latest GenRTX A6000, L40, A100 80GB, H100 PCIe
Multi-GPU NodesDeploy anything from a single GPU instance to 8x GPU servers for training, inference, and batch workloads.
PCIe vs SXMH100 PCIe

Pricing Model

Per MinuteBilled per minute with expandable configurations for any project
SubscriptionLow-cost cloud GPUs, on demand
Public Pricinghttps://www.thundercompute.com/pricing
Egress ChargesNo, Thunder Compute does not bill for data egress.
Pay-as-you-goPay only for what you actually use.

Performance & Scaling

Multi-Node TrainingDeploy anything from a single GPU instance to 8x GPU servers for training, inference, and batch workloads.
Elastic ScalingBilled per minute with expandable configurations for any project

Developer Experience

OnboardingA Thunder Compute account; authentication via OAuth in your browser; no API tokens or environment variables required.
FrameworksPyTorch, CUDA
CLI ToolingCLI examples and integrations: claude mcp add; codex mcp add; opencode mcp add / mcp auth; npx @smithery/cli install; curl examples.
Jupytersetup-jupyter — Launch a Jupyter Lab environment on a GPU instance
TemplatesOS templates (Ubuntu, PyTorch, etc.)
DocumentationDocumentation Index; MCP Server guide with tools and prompts; tool reference tables for Instance Management, Information, Snapshots, SSH Keys, Port Forwarding, Connectivity, Billing & Usage, API Tokens; Example Usage; Troubleshooting; MCP Directories.
API FeaturesCLI port manipulation tools, RESTful API for instance management

Security & Compliance

SecurityAutomatic HTTPS,DDoS protection
ComplianceAutomatic HTTPS, DDoS protection
Listed in MCP directories (SmitheryMCP RegistryGlamaPulseMCP)references to Terms & ConditionsPrivacy Policyintegrations with AI agents Claude CodeCursorWindsurf, and Codex.

Data Center Locations

Coverage

Compliance Regions

Datacenter Locations

Key Strengths

Low-cost transparent per-minute GPU pricing
fast provisioning
persistent snapshots
enterprise-grade data centers
proprietary virtualization to improve scheduling efficiency
no data egress fees
MCP server integration for AI agents.

Additional Information

Support Options

["Integrated developer tools via VS Code","CLI for port forwarding and tunneling","Comprehensive documentation"]

Community

Smithery listing (one-click install), MCP Registry, Glama, PulseMCP

Core Proposition

Low-cost cloud GPUs, on demand

Last updated March 2026. Information subject to change.