Model ReleasesLiquid AI's new 24B MoE model delivers massive reasoning power with only 2B active parameters, optimized for high-throughput local AI agents.hace 3 días6 min de lectura
