Cline's Pareto 26.10 Preview routes each prompt to several frontier and open models, grades their answers, and returns the best one while preserving prompt cache. Cline says it matches Fable's DeepSWE score at $0.24 per task versus $13.41, about 56 times cheaper.
Pareto 26.10 Preview is a new model available in Cline.
It routes to several frontier and open models, grades the answer, & returns the best one - while preserving prompt cache + keeping cost low.
~56x cheaper than Fable for the same DeepSWE score at $0.24 per task vs $13.41!
Source: Cline · x.comPublished · added here
