Compare K2 Horizon 0.9B with measured models

A reliable intelligence-versus-cost verdict requires both models' intelligence scores and benchmark-run costs. Missing data is not a performance verdict.

Measured matchups