In design — not operational

HiveCredits

The economic layer we're designing to compensate compute contributors and model publishers. Currently a design, not a running system.

Where HiveCredits stands today

HiveCredits is the working name for the unit of value that will mediate compute and model usage across ColabHive once the contributor program is operational.

Not operational. No tokens, no marketplace, no settlement happening today.
No public price. The exchange rate, fee structure, and contributor splits are still being designed and will not be published until they're real.
Not a token. HiveCredits is an internal accounting unit, not a cryptocurrency. No blockchain, no airdrop, no presale.

What it's intended to do

Once the program is live, HiveCredits should:

Account for compute that contributors provide on their hardware
Account for usage of public agents/models that builders publish
Mediate borrowing and lending of capacity between contributors when local hardware is over- or under-utilised
Settle these flows through an escrow-and-ledger layer (the "Financial Core" in the architecture)

None of this is running in production today. All compute is currently provided by our anchor customer across three sites in LatAm.

Why this page is short. Until HiveCredits has actual flows, real settlement, and a documented economic model, anything more detailed would be selling something that doesn't exist yet. The architecture supports it; the program is being designed; the page will get fleshed out when it's real.

Architecture context

The architecture diagrams on the home page describe how Public Pool, Private Pool, escrow and settlement are designed to interact. Those flows are design targets, not the current operating mode.

The IEEE-format paper at the research section describes the broader energy-aware orchestration model.

Want to discuss?

If you have GPU capacity to contribute, models to publish, or just want to follow this work, get in touch.

[email protected]