The control layer for your AI.

Who it's for

AI power users already on two or more AI providers, clients, or runtimes who want one desktop control layer to coordinate them for real work.

Why it exists

AI adoption fragments across provider, client, account, and tooling at once. AIObox's plan is to unify those axes behind one control layer: choose the AI, combine them, execute locally, automate the recurring work.

Built on what already works

AIObox's core economics are already proven by a separate, shipping project: aki-mcp-sv lets Claude, ChatGPT, Grok, and Gemini read, edit, and run commands on a local machine using web quota instead of per-token API billing, with multi-account access via browser profiles. AIObox's plan is to package that same model into an app with no command-line setup, across more providers at once. That packaging is not built yet.

Rule sync, already running

Keeping instructions consistent across every connected AI account is not new engineering risk either: it generalizes AkiDevRule, an open-source shared rule corpus already installed with one command across five different AI coding tools. AIObox's plan is to automate that install per account instead of running it by hand.

Where things stand today

The only shipped part of AIObox is this web informational surface. See pricing for the Energy wallet plan and the docs for more on what's live and what's planned.

Ownership

Owned and managed by akitao.com.