Solana-native optimistic oracle

STAKE YOUR TRUTH

Opal resolves assertions through a default optimistic answer, a first LLM resolution round, and a final private voting escalation when the LLM result is disputed.

Resolution advances in three layers

Each layer is economical and terminal settlement is only available once the assertion is fully resolved.

01Assert

Post assertion & bond. Default answer: True.

02Dispute

First dispute opens the LLM round.

03Escalate

Second dispute opens private voting; settlement on Resolved.

State determines the current answer

Asserted

State

Default answer is true. The statement can still be challenged.

AssertedLLM

State

The first dispute resolved. Consumers read LlmResolutionRound.outcome.

PendingVote / Voting

State

The LLM result is under challenge and the final answer is not settled yet.

Resolved

State

Outcome is final, settlement is irreversible, and integrations can safely settle.

Built for integrators, challengers, and protocol operators

Prediction markets

Opal

Settle subjective markets only when the assertion reaches Resolved. Consumers can read the state machine and wait for the terminal outcome.

Disputers

Opal

Challenge optimistic defaults or challenged LLM outputs for direct economic upside. The protocol is designed to reward accurate disputes.

Builders

Opal

One assertion PDA can expose the current state, round pointers, dispute count, and final answer for downstream applications.

Future operators

Opal

Switchboard, Nosana, and MagicBlock integrations are reserved in the accounts, but v1 does not require those systems to be live.