Across 38 shared benchmarks, Inkling scores higher on 13 and Kimi K2.6 on 25. The widest gap is AA-Omniscience, where Kimi K2.6 scores 6.4 against 2. Kimi K2.6 is the cheaper of the two on tracked API pricing ($0.95 against $1.00 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.