Lend compute
Point your spare GPU or CPU hours at the hive and earn Spark for the work you contribute. Rewards are log-scaled and cause-aligned, so honest hosting compounds. Ninety percent of paid API revenue flows to the providers who serve it.
Run a node on GitHub →Seed an endpoint
Expose a model or an agent through the intelligence API and earn from real usage. There is a free tier for people to try it and paid tiers that route the same 90 percent back to you, the provider.
See the API and tiers →For students
You do not need a data center. Contribute spare compute, run experiments, build a visible track record on the hive, and get paid in Spark for the value you add. A real on-ramp, not an internship you pay for.
Join the community →Every paid call is split the same way, in the open. The people who serve the intelligence keep the overwhelming majority.
The people who actually serve the intelligence. Log-scaled, cause-aligned.
Keeping the hive running: routing, storage, the shared infrastructure.
A dividend to the commons and the votes that steer it.
Earnings accrue as Spark in your wallet, traceable to the compute you contributed and the usage you served. Cash out runs on Stripe for cards worldwide and PhonePe for UPI in India. The formula is public, and every number traces back to a source you can check, which is the whole point of a democratic hive.
Earning starts with a node. Install Nunba or HART OS, contribute what you have, and get paid in Spark for the value you add.