MoRe matches multi-agent work at 20x less cost

Mixture of Roles (MoRe) reports multi-agent-level performance at 20× lower token cost by composing multiple learned role vectors into one single-turn inference path; it also beat single-agent baselines by 2.2% on average. The result is a research benchmark, not a production guarantee, but builders testing multi-agent orchestration should add a single-call role-composition baseline beside their current swarm before paying for more context and coordination, as the archive’s multi-agent coordination analysis also suggests.