A marketplace built on Pearl's useful-work mining: pool your hashrate today, sell LLM inference from the pooled fleet next, trade mining yield after that. One stack, honest status labels, no invented numbers.
Each track is labeled with its real status. We ship the pool first, the inference market second, the yield market third.
Connect any PRL miner over standard stratum and earn PPLNS payouts. Our unique angle: we are building the only pool that can connect the official vLLM dual-use miner — the H100/H200 miner that produces PRL as a by-product of LLM inference and is solo-only everywhere else. Standard stratum mining is live; the dual-use connector is in betaBeta.
Connect your rig →Buy LLM inference served on the pooled GPU fleet through an OpenAI-compatible API. GPU contributors get a share of the inference revenue on top of the PRL they mine. A credits system with Stripe billing is already built; the market opens when the fleet does.
Get notified →Buy PRL mining yield without owning hardware, or — if you run a data center — swap your mining yield for compute. NiceHash-style order matching on top of the pool. This track is planned, not built; we publish it so you can see where the marketplace is headed.
Register interest →Pearl's proof-of-work is a noised matrix multiplication — the same math that runs every LLM. That is what makes a mining pool, an inference market and a yield market one product instead of three. Read how it works →