Artificial Analysis(@ArtificialAnlys):Claude Opus 5's effort setting spans a wide range of token usage-performance tradeoffs. On GDPval-AA v2, effort levels span 407 Elo points, with output token usage ranging around 8x from low to max effort. Like with GPT-5.6 Sol, this means Opus 5 can use either far fewer or far more tokens to complete the evaluation than models from other labs, depending on effort settings

Artificial Analysis

@ArtificialAnlys

Independent analysis of AI

加入 January 2024

664 正在關注 127.6K 粉絲

Artificial Analysis@ArtificialAnlys

2026.07.24 19:18

Claude Opus 5's effort setting spans a wide range of token usage-performance tradeoffs. On GDPval-AA v2, effort levels span 407 Elo points, with output token usage ranging around 8x from low to max effort. Like with GPT-5.6 Sol, this means Opus 5 can use either far fewer or far more tokens to complete the evaluation than models from other labs, depending on effort settings