Gemini CLI writes it. The grid runs it.

Connect once and every project Gemini builds with you can go live at a real URL, with the database, memory, and AI already there.

Make your creation live See how it works

Get connected

Connect once. Keep building.

EasiestCopy and send this to Gemini CLI

Gemini can install it for you. Nothing to type in the terminal.

Install the CloudGrid MCP server for me.
Run this for me:
gemini mcp add cloudgrid npx -y @cloudgrid-io/mcp
Tell me when CloudGrid is connected, then ask me what to build.

OrDo it yourself

Paste this in your terminal. Gemini saves it for you.

gemini mcp add cloudgrid npx -y @cloudgrid-io/mcp

Ready when you need it

Built in, out of the box

Tell Gemini CLI what your project needs. The Grid connects the services that make it work.

AI, data, files, and a runtime that stays live, all together on your grid.

Connected

AI, no keysYour creations call models through the grid. No key wrangling, no billing setup.Ready
DatabaseAsk for one and it exists, already connected to what you are building.Ready
Cache and storageFast paths and files, wired in before you think to ask.Ready
Stays liveRuntimes are watched, versioned, and kept alive. Logs and rollback included.Ready

Good to know

A few questions.

Does CloudGrid work with Gemini CLI?

Yes. Add the CloudGrid MCP server or install the CLI and skills, then tell Gemini to plug your project in. Seconds later it lives at a real URL.

Do I need to leave the terminal?

No. Gemini runs the whole thing: build, plug, and hand back the live URL.

What happens to the project after it goes live?

It lands on the grid with a database, memory, and AI behind it, where it can be found, shared, and picked up.

Keep building in Gemini CLI. Let the grid take it live.

Connect once, then make your next creation live.

Request an invite

Already working with an agent?

fetch https://cloudgrid.io/agents and plug this in