Models bigger than your PC
Your verified work earns credits. You spend them on models no single gaming card can hold — the top one runs across a Chain of home GPUs.
Frontier-level AI, for everyone.
Lend your PC when you don't use it. In return, use AI models that are too big for your PC — split across many home GPUs, like one big machine.
How it works
-
01
Install the app
A small open-source app runs on your PC. You choose when it works: only when you are not gaming, only at night, or whenever the PC is on anyway.
-
02
Your GPU joins the network
It answers requests on its own, or becomes one stage of a Chain: a large model split across several GPUs. Every answer is checked, so honest work counts and cheating doesn't.
-
03
You earn access
Verified work earns you credits. You spend them on models your card cannot run alone → launch models
The Chain
A big model doesn't fit on a gaming card. So we cut it into slices.
Each GPU holds a few dozen layers. A request flows through the cards in order, like an assembly line, and many conversations share the line at the same time. The full model never needs to sit on one machine.
We start with short chains of three or four cards in the same region, and lengthen them only when needed.
- modelQwen3.8 Flash Next
- quantizationIQ3_XXS · 43.8 GiB
- chain3/4 GPUs
- throughput160–180 tok/s
- networkRTT ≤ 25 ms · ≥ 100 Mbit/s
What you get
A free quota to start
Everyone gets a free daily quota in the chat, so you can see it work before you lend anything.
Credits
Credits are the network's internal unit. Your GPU earns them for every answer it computes and that passes verification: each model has one price list, and what users spend is exactly what the GPUs earn — split across the cards of a Chain by the share of the model each one carries. You spend them on any model, in the chat.
WIP
converting credits to $$$
not promised: only if and when the network pays for itself
What you give
YOUR GPU
An NVIDIA card, 8 GB or more
8–12 GB cards run the base model on their own. 16–24 GB cards also take a slice of the top model in a Chain. Windows or Linux.
YOUR TIME
The hours you choose
Only when your PC is idle, only at night, or whenever it's on anyway. Your PC uses electricity while it works: we size credits so the access you get is worth at least three times that cost. Offered hours are not paid hours: verified work counts most.
YOUR LINE
A decent home connection
For the Chain: about 100 Mbit/s up and down and under 25 ms to the nearest servers. Chains are built inside one region, so the cards stay close.
Why it pays off
Your card is already paid for. What it costs to run is electricity — and that buys a lot more AI than a subscription.
PRICE PER TOKEN
0.04–0.12 €electricity per million tokens of the 26B model on a 16–24 GB card
0.22–0.40 $what APIs charge per million tokens of the same class
The medium model costs 2 to 8 times less on a home GPU than on the market.
HOURS, NOT DOLLARS
~170 Mtokens a month from a 16 GB card running 8 hours a day
~1.5 Mtokens a month used by a heavy chat user
A few hours of your card cover your own use. The rest serves the network and earns you access to bigger models.
THE TOP MODEL
≈ 1×a Chain's electricity vs the market price of Qwen3.8 Flash Next
On the biggest model we don't beat data centers on price — we're about even. The win is access: no gaming PC can run it alone, and you pay for it with idle hours instead of a monthly bill. With solar power, or a PC that's on anyway, it costs you close to nothing.
Our estimates, November 2026: electricity at 0.314 €/kWh; energy per token measured by us on rented GPUs; market prices from public price lists. They will change, and so will these numbers.
Launch models
Swipe through the roster we plan to launch with. Tell us what you want next.
BASE · FREE PLAN
Gemma 4 12B
Everyday chat and summaries. Runs on a single card.
- to contribute
- 8 GB+
- quantization
- Q3
MEDIUM
Gemma 4 26B-A4B
Stronger reasoning, code and long documents. One 16 GB card, or a Chain of smaller ones.
- to contribute
- 16 GB+
- quantization
- IQ4_XS
TOP · IN THE CHAIN
Qwen3.8 Flash Next
A frontier-class open model for coding, split across 3–4 home GPUs. LiveCodeBench v6: 86.3 in this quantization (publisher's figure).
- to contribute
- 16–24 GB
- quantization
- IQ3_XXS
Subject to licence confirmation for hosted use.
NEXT?
Want a bigger model?
Another model, another quantization, a bigger one? Vote on the roster →
Open source where it matters
You run our code on your machine. You should be able to read it.
The host app and the Chain stage — everything you install — are open source, together with the engine patches they build on (a fork of llama.cpp). Read it, build it, audit it.
$ publicompute-host --start --only-when-idle › gpu: RTX 4070 Ti Super · 16 GB · ready # source: github.com/publicomputeView the code on GitHub
Shape the roster
Which models should the network run next? It takes 30 seconds.
Pre-register your rig
Self-declared, no install needed. It helps us know which PCs the network will have at launch — and which Chains we can build.
Questions
Will it slow down my games?
No: the app works only in the hours you allow, and stops when you start playing.
Can other people see what my GPU processes?
Your PC processes slices of a model's computation, not files of yours. We never access your files. Privacy options for users of the network are described at launch.
How do you stop cheaters?
A share of every node's work is recomputed by independent verifiers. In our tests honest nodes were never rejected and degraded models were always caught.
Is this crypto?
No. No coin, no speculation. Credits are an internal unit to use the network.
When do you launch?
When the network is ready and enough people have joined the waitlist. Pre-registered rigs are invited first.