Your Gaming PC Could Earn Money Running AI Jobs While You Sleep

Key points

• Companies including Evolving Edge, Far Labs, Salad, and Bless Network pay home computer owners to run AI inference, the process of feeding a question to a pre-trained AI model and getting an answer.
• Evolving Edge is in open beta; Far Labs plans to launch its platform, Far AI, within weeks.
• AI inference on consumer hardware costs less partly because these companies carry no data-centre construction costs.
• Security protections include open-source scheduling software, encrypted communications, and a "least privilege" design that limits what either side can see.
• Distributed networks of small devices could also reduce outages of the kind that leave single data centres as one point of failure.

Your laptop might sit idle for fourteen hours a day. A gaming PC gets a few hours of use at night. A home server in a basement closet never quite reaches its potential. Now, a cluster of start-ups wants to pay you to put all of that dormant hardware to work.

The pitch is simple: AI companies need computing power to run AI inference, which is when a ready-built model processes a question and produces an answer. Until recently, that work happened almost exclusively inside giant data centres. These new platforms want to spread it across millions of ordinary devices instead.

"Imagine Uber or Airbnb, but for AI inference computing tasks," says Ilman Shazhaev, founder and CEO of Far Labs, based in Abu Dhabi.

How does it actually work?

Signing up is meant to be straightforward. Install an app, set a schedule for when the platform can use your machine, and your device joins a shared pool of computing resources. Jobs arrive, your hardware processes a slice of the work, and you get paid.

John Federico, Evolving Edge's founder and CEO in Austin, Texas, has run home servers for years. The idea struck him when he heard the constant talk of an AI compute shortage. "92 percent of the country has broadband, and you have people like me who have mini data centres in a closet," he says.

Security is the obvious concern. Evolving Edge open-sourced its scheduling software so anyone can inspect exactly what it does to their machine. Far Labs built its system around a principle called "least privilege": the host and the company running the job each get access only to the minimum information needed to complete the task. Inference jobs run in an isolated environment with encrypted communications, and hosts can pause or remove the software at any point.

What about models that are too big for one computer?

Smaller, open-source AI models often fit on a single consumer GPU, the specialised chip that handles heavy number-crunching. When a model is too large, the platforms split the workload.

Evolving Edge uses an open-source tool called Ray to divide an inference task across several devices. Far Labs built its own software that cuts the model into pieces, ships those pieces to different machines, and reassembles the answers through a central coordinator.

This isn't just a cost play. Shazhaev argues that a distributed network is also more dependable than a single data centre. "To shut down Bitcoin, you need to nuke the whole planet. Here, we have the same concept," he says.

Far Labs claims its network delivers responses in 100 milliseconds or less, fast enough for real-time uses like in-game AI video generation.

Should you worry about what runs on your machine?

These questions aren't hypothetical. The same compute-shortage pressures driving this trend are ones we've tracked across several recent stories, including how French startup Kog claims software alone can speed up inference on chips businesses already own. That story from 14 August 2026 explored the software side; this one lands the costs in your living room.

What should potential hosts watch for?

Before downloading anything, check three things. Confirm the scheduling software is genuinely open source and auditable; an honest platform points you straight to the code. Read what the app can access, looking for explicit limits on CPU, GPU, memory, and storage usage, with no broad system permissions. Verify you can pause or remove the software at any time without forfeiting earnings already accumulated.

These platforms are early. Evolving Edge is in beta; Far Labs hasn't launched yet. Treat both as experiments, not reliable income. The concept has history: SETI@Home ran from 1999 to 2020, using volunteer computers to scan radio telescope data for signs of extraterrestrial life. Commercial versions of the same idea carry more scrutiny of your hardware, so the due diligence matters more.

The real thing to watch isn't the earnings. It's whether these platforms can hold security guarantees at scale, across thousands of devices they don't own.