Vals AI Says Claude Opus 5.5 Keeps 80% of Score at 16% Cost

Vals AI reports that four of seven frontier models kept at least 80% of their Vals Index score at half the cost or less when run at their lowest reasoning setting. Claude Opus 5.5 held 80% of its score at 16% of the cost, researcher Derrick Liang writes. In 22 of 56 model-benchmark pairs, cheaper settings matched leaderboard scores.

• Gemini 4 Argon holds 99% of its Vals Index score at 84% of the cost, Derrick Liang reports.
• Legal Research shows a median 19.2 point score gain from the lowest to highest reasoning setting, Derrick Liang reports.
• On the Vals Index, none of the seven models matches its leaderboard score at the lowest setting, Derrick Liang reports.