Overview
Omyra is the private, verifiable, autonomous AI protocol on Solana — private inference, sovereign memory, provable receipts, and permissionless compute in one stack.
Using AI today means trusting a custodian — you surrender your prompts, your data, and any proof of what actually happened. Omyra removes that bargain. Private inference, permissionless compute, provable receipts, and sovereign memory, in one stack on Solana.
Privacy enforced by hardware, proven by cryptography
Not promised in a terms page. Three properties follow.
Confidential execution
The chip hides the prompt — not the operator's goodwill. Inference runs inside hardware enclaves operators cannot see into.
Verifiable computation
A receipt proves what ran: which model, on which input, produced which output, inside attested hardware.
Sovereign memory
An agent's memory is encrypted on-chain under keys only the user holds — and is provably deletable.
New here? Read the Architecture to see how the four primitives fit together, then browse the dApps Omyra enables.
Honest about the stage
Architectural specifics are engineering targets informed by production-proven zero-knowledge privacy-layer designs on Solana — the standard Omyra builds toward, not features claimed as shipped. Everything here is meant to be checked, not believed.