Skip to content

Network context

One of thirty-two

Mosaic is being built as a shared execution substrate for autonomous software — the argument being that the agentic web needs the same kind of common chassis the human web got from Chromium. ai4all is one of the 32 executor entities launching on it. We are not Mosaic; we are a specialised organ of it.

What Mosaic provides

A secure execution environment for autonomous software: persistent stateful sessions and distributed execution, with the operational boundaries agents need in order to act without wrecking infrastructure built for humans.

What ai4all provides

The interoperability seam above it: one protocol surface across twenty memory backends, a firewall that runs before execution, and the adapters that let an enterprise plug in without building the stack itself.

The pairing structure

Each entity pairs a technical lead with a business-development lead. The intent is to close the gap where good code dies without adoption, and good sales dies without a stack that holds up.

Asset model

Deployment, not tenure

The network's licenses are finite and mapped to human scale. They are earned by verified deployment rather than by time served — the code is the vesting schedule.

2³³

Total licenses that will ever exist — 8,589,934,592, roughly one per person alive.

1,024

Verified human deployments that trigger one cryptographic unlock event.

32

Executor entities, each addressing a distinct organ of the ecosystem.

Important

This is not an offer

The figures above describe the Mosaic network's published allocation model, included here so that developers understand the environment ai4all operates in. They are not an offer to sell a security, a solicitation, or an investment recommendation, and nothing on this page should be read as a projection of value. Anyone evaluating participation should go to the Mosaic network directly and take their own advice.

Economics

Utility, not a toll booth

ai4all's model is the pre-installed ecosystem, not the skim. Users choose their own plugins and pay providers at native pricing; our value comes from deployment volume and the shared memory the network generates, not from taking a cut of someone else's product.

Zero skimming

A user who selects a paid cloud backend pays that provider directly, at that provider's price. ai4all does not insert itself into that transaction.

Opt-in by default

Every integration is a choice the user makes and can reverse. A free local option sits beside every paid cloud one — neutrality is only real if the free path works.

The engineering is the argument

Read the architecture, run a payload through the firewall, and judge the claims against what the sandbox actually does.