Microsoft

BenchLM aggregates canonical model families under Microsoft so you can compare release depth and model quality at the provider level instead of jumping across single model pages. Scores shown here are provisional unless otherwise noted.