Across 32 shared benchmarks, Kimi K2.6 scores higher on 31 and Kimi K2 Instruct on 1. The widest gap is browsecomp, where Kimi K2.6 scores 86.3 against 14.1. Kimi K2 Instruct is the cheaper of the two on tracked API pricing ($0.60 against $0.95 per million input tokens).
Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing. Quoted rates are the price-setter row we currently track for each model — its direct or vendor-official listing where one exists (direct, direct), otherwise the lowest tracked offer.