Benchmarks / Composite
Arena Text
Crowdsourced pairwise human preference on open-ended text prompts (style control on); Bradley-Terry rating.
Arena Text
Higher is better · best effort per model
full table below
Benchmarks / Composite
Crowdsourced pairwise human preference on open-ended text prompts (style control on); Bradley-Terry rating.
Higher is better · best effort per model
full table below