Savings Plans are now available for on-demand compute resources.

Users can now maximize savings with discounts for long-term monthly spend commitments:

• 12-month term:** 30% discount**
• 36-month term: 50% discount

Here's how it works

You can now commit to a specific monthly expenditure for compute resources, prepay for the desired duration, and receive discounts based on the commitment length.

Each month, the committed amount offsets the on-demand compute usage on your project bill.

Head to Settings> Savings Plans to create your first plan, or check out our Docs to learn more.

We've updated our pricing for reserved instances, and** bare metal GPU** options now have increased discounts:

• Monthly: 30% off the hourly rate
• Yearly: 50% off the hourly rate

Learn more

Introducing the g4.rtx6kpro.large

The newest addition to our bare metal GPU lineup, this instance is powered by NVIDIA’s brand-new RTX PRO 6000 Blackwell GPUs.

Here's the complete spec:

• **GPU: **8x NVIDIA RTX PRO 6000 Server Edition (96GB per GPU)
• CPU: 2x AMD EPYC 9355 (64c @ 3.55 GHz)
• RAM: 1.5 TB of DDR5
• STR: 4x 3.8TB NVMe
• NIC: 2x 100 Gbps

Starting at just $1.97 / gpu / hr, the g4.rtx6kpro.large is now available for capacity reservation at selected locations.

Learn more

You can now connect AI assistants and other developer tools directly to your Latitude.sh infrastructure, enabling natural language control of your bare-metal servers and platform resources.

Our MCP server provides:

• Instant AI integration via npx or drag-and-drop DXT installation for Claude, Cursor, VS Code, and Windsurf
• Natural language operations to list, create, and manage servers, projects, and SSH keys through chat
• Simple authentication with just your Latitude.sh API key - no complex configuration required
• Full API exposure as AI-ready tools, bringing infrastructure automation directly into your IDE conversations

Check out the MCP docs to learn more

Developers can now interact with Latitude.sh's API natively in TypeScript or JavaScript, with full type safety and async support. The SDK streamlines infrastructure automation with:

• Quick setup via npm/yarn with built-in API authentication and TypeScript definitions
• Complete API coverage for servers, projects, virtual networks, firewalls, storage, and more
• Production-ready features, including error handling, automatic retries, and pagination
• Modern JavaScript support with async/await, tree-shaking, and both CommonJS/ESM compatibility

Check out the SDK docs to learn more