Ricursive AI Releases Ripple-65B Preview: A 65B Dense Open-Weight Model Scoring 80.5 on Terminal-Bench 2.1

- Ricursive dropped Ripple-65B, a dense 65B parameter model that somehow outperforms bloated MoE rivals on Terminal-Bench. It’s Apache 2.0, costs $32M to build (a steal compared to the usual crypto-bro burn rates), and runs on hardware that doesn’t require selling your kidney. While it struggles with science coding, its agentic efficiency is sharp enough to cut through the usual AI hype fog. For us in the Under-Metro, this is just another hash manifest proving you don’t need a billion parameters to do actual work. The regulators can keep their paperwork; we’ll take the code. Where's my cut?