Distributed AI Compute Marketplace

MacMesh Your Mac powers the world's AI — while you sleep.

Every Apple Silicon Mac is a dormant supercomputer.
We wake them up. Together.

Explore
The Imbalance

One side pays too much.
The other side sits idle.

Cloud GPU compute is a luxury. Startups bleed money renting what they need. Meanwhile, millions of the most powerful personal computers ever built — do nothing for 80% of their lives.

Cloud GPU
$2–5/hr
Per task. Per hour. Forever. The toll gate that keeps AI out of reach for individuals and small teams.
VS
Mac GPU Utilization
~20%
The remaining 80% is wasted potential. Apple Silicon is optimized for AI. It just needs a reason to run.
The Concept
What if every idle Mac became a node
in a global AI compute network?
I
Connect
Mac owners join the network. Their machines become available when idle. Zero effort.
II
Compute
AI workloads are distributed across the mesh. The right task finds the right machine, automatically.
III
Earn
Processing completes. Rewards accumulate. Your Mac pays for itself while you sleep.
The Connection

The infrastructure layer
behind the Obra ecosystem.

MacMesh doesn't exist in isolation. It's the engine room that powers the most demanding workloads across the Obra platform — and anyone else who needs distributed GPU compute.

EZRA
Video generation requires GPU power. MacMesh provides it — distributed, not centralized.
DaviCore
LLM inference at scale. Multiple models, multiple nodes — no single point of failure.
The World
An open marketplace. Any AI workload, any client — not just Obra.
The Scale

A sleeping giant,
waiting to be woken.

200K+
Mac Studios sold per year.
Each one, a potential node.
$4.5B
Distributed compute market.
Growing 28% annually.
128GB
Unified memory per node.
Purpose-built for AI workloads.
2,000 nodes × 128GB unified memory = a cluster rivaling AWS GPU instances
The world's most powerful personal computers
are sitting on desks, doing nothing.

We're building the network
that turns them into an army.
Read the Obra manifesto →