Latitude.sh
Bare metal servers and virtual machines deployed globally, managed through a dashboard, API, CLI, SDKs and Terraform.
At a Glance
Engagement
Available On
Alternatives
Listed Oct 2026
About Latitude.sh
Latitude.sh is an infrastructure platform for deploying and managing bare metal servers and virtual machines globally. Its documentation describes combining cloud flexibility with the performance, security and control of dedicated infrastructure. Workloads can be run from the dashboard, API, CLI, SDKs and Terraform provider.
What It Is
Latitude.sh is a cloud and dedicated-infrastructure provider. Users sign up, add a payment method and verify their account, create a project to group servers and network resources, then pick a plan, location and operating system to deploy a server. Compute options include dedicated single-tenant bare metal (including Metal GPU plans), general-purpose virtual machines with backups, restore and monitoring, and LKS Preview, a managed Kubernetes service on bare metal.
Storage and Networking
The platform offers S3-compatible object storage, block storage (private preview), shared file storage and managed PostgreSQL with backups and monitoring. Networking covers private Layer 2 networks, public routed IP blocks, Elastic IPs, BYOIP, resource-level firewalls and DDoS protection that the docs say is built and operated by Latitude.sh. Cloud Gateway and Global Gateway provide private connections to major cloud providers and between Latitude.sh locations.
Developer Tooling and AI Agents
The API follows JSON:API and covers every platform resource. The lsh CLI is open source, and official TypeScript, Go and Python SDKs are available, along with a Terraform provider and a Postman collection. A hosted MCP server lets AI agents manage infrastructure. A public examples repository demonstrates custom OS images over iPXE, cloud-init user data and Terraform deployments.
Account Management
Projects are invoiced separately, and teams can invite members and manage roles. Account security options include MFA, SAML SSO and passkeys.
Community Discussions
Be the first to start a conversation about Latitude.sh
Share your experience with Latitude.sh, ask questions, or help others learn from your insights.
Pricing
Metal GPU g4.b300.large
Dedicated GPU server with 8 x NVIDIA HGX B300, billed hourly or reserved monthly or yearly
- 8 x NVIDIA HGX B300, 800Gbps Dual Plane RoCE
- 2304 GB VRAM (288 GB per GPU)
- 2 x 960 GB NVME, 4 x 7.6 TB NVME
- 20 TB free out / ∞ in
- $64/hr
- $46,720/mo
- Reserved: monthly 50% cheaper than hourly, paid yearly 65% cheaper than hourly
Metal GPU g4.rtx6kpro.large
Dedicated GPU server with 8 x NVIDIA RTX PRO 6000, billed hourly or reserved monthly or yearly
- 8 x NVIDIA RTX PRO 6000
- 768 GB VRAM (96 GB per GPU)
- 4 x 3.8 TB NVME
- 20 TB free out / ∞ in
- $24/hr
- $17,520/mo
Metal GPU g3.h100.small
Dedicated GPU server with 1 x NVIDIA H100 80GB, billed hourly or reserved monthly or yearly
- 1 x NVIDIA H100 80GB
- 80 GB VRAM
- 2 x 3.8 TB NVME
- 20 TB free out / ∞ in
- $1.68/hr
- $1,230/mo
Metal rs4.metal.xlarge
Bare metal server, billed hourly or reserved monthly or yearly
- AMD 9554P, 64 Cores @ 3.1 GHz
- 1536 GB RAM
- 2 x 480 GB NVME, 4 x 8 TB NVME
- 2 x 100 Gbps network, 20 TB free out / ∞ in
- $5.44/hr
- $3,971/mo
Virtual machines vm.small
General-purpose virtual machine, billed hourly or reserved monthly or yearly
- 4 vCPUs
- 16 GB RAM
- 160 GB Local storage
- 20 TB free out / ∞ in, 10 Gbps
- $0.09/hr
- $69/mo
Latitude for Enterprise
Bespoke bare metal deployments; contact sales for custom pricing
- Deploy one or a thousand fully automated bare metal servers specific to your needs
- Custom pricing and volume discounts
Capabilities
Key Features
- Bare metal servers including GPU plans
- Virtual machines with backups, restore and monitoring
- Managed Kubernetes (LKS Preview)
- S3-compatible object storage
- Block and file storage
- Managed PostgreSQL
- Private and public networks
- Elastic IPs and BYOIP
- Firewall and DDoS protection
- Cloud Gateway and Global Gateway interconnection
- Hosted MCP server for AI agents
- MFA, SAML SSO and passkeys
- Team management and per-project invoicing
