jpval-deepseek · status: completed

Bench results

capability fixture model status score cost($) dur(s)
refresh jp-8 deepseek-v4-pro success 7.25 0.090096 1,920.4

Winners

  • score: deepseek-v4-pro
  • cost: deepseek-v4-pro
  • speed: deepseek-v4-pro
  • overall: deepseek-v4-pro