Small doesn’t mean less capable. We trained a small model using @reasonblocksinc for WorkBench: Opus-level accuracy at ~1/132 of Opus’s measured inference cost. All our charted models sit on the cost–accuracy Pareto frontier. Need your own trained small model? DM me. Read
Small Model Matches Opus Accuracy at 1/132 Inference Cost
A developer trained a small model with reasonblocksinc for WorkBench, claiming Opus-level accuracy at roughly 1/132 of Opus's inference cost, sitting on the cost-accuracy Pareto frontier.