Compute that already exists, put to work.
SpiderMesh runs Verity's model work across machines people already own, instead of racks rented by the hour. That is what makes the cheap tiers cheap, and it is why the work adds no new cooling load.
Contribute, place, verify.
A machine offers the capacity it is not using. Verity places a bounded unit of work on it, and on a second node owned by somebody else. Both return a digest. The result counts only when the two digests match, so a single dishonest or broken node cannot decide an answer on its own.
The price follows spare capacity.
A datacentre charges for a machine bought, powered and cooled for you. SpiderMesh pays for time on a machine that already exists and is already on. Contribution is denominated in node-hours rather than currency: one average node contributing for one hour earns one credit.
No new cooling load.
Large datacentres evaporate water to stay cool, which is why they are built where water is. A node in somebody's room is already cooled by the room. SpiderMesh adds work to machines that are running anyway rather than adding buildings that have to be kept cold.
What this build does not do yet.
Node enrolment and workload submission are not available from this website. The network, the coordinator and the consensus rule are in the source and running; the controls that would let you enrol your own machine from a browser are not part of this build. No earnings, capacity or uptime commitment is made here.
Read the technical record.
The network design, the consensus rule, the credit ledger and the gateway delivery path are documented with the source.