Across 14 shared benchmarks, Agnes 2.5 Pro Beta scores higher on 10 and K2 Horizon 375B A23B on 4. The widest gap is critpt, where Agnes 2.5 Pro Beta scores 15.7 against 4.6.
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.