Skip to content
Trending storyDeveloping

Higher effort settings lift HLE-Diamond scores for Claude Opus 5.5 and GPT-6 Astra at higher cost

1 article1 sourcesince Oct 29Last article 1h ago ·

Overview

AISummary of 1 article

Artificial Analysis reports that HLE-Diamond scores rise as effort settings increase for Claude Opus 5.5 and GPT-6 Astra, and that per-task cost rises with them.

Claude Opus 5.5 goes from 42% at low effort to 68% at max, a 26-point gain at about 35 times the cost per task. GPT-6 Astra goes from 45% to 64%, a 19-point gain at 11 times the cost per task.

Written by AI from the articles below · updated Oct 29, 10:14 PM ET

Check the sources:

Latest developmentOct 29, 10:09 PM ET
Higher effort lifts HLE-Diamond scores but raises per-task cost sharply

Article timeline

The articles in this story. Times are ET.

Oct 29
  1. Artificial AnalysisOfficial
    Higher effort lifts HLE-Diamond scores but raises per-task cost sharply

    AIArtificial Analysis reports that HLE-Diamond scores rise with higher effort settings, but cost rises too. Claude Opus 5.5 goes from 42% at low effort to 68% at max, a 26-point gain at about 35x the cost per task. GPT-6 Astra goes from 45% to 64%, a 19-point gain at 11x the cost per task.

    Image from @ArtificialAnlys's post

Heat Trend

Not enough continuous observations to show a trend yet.