Skip to content
Trending storyDeveloping

Artificial Analysis makes Time per Task its headline API speed metric

1 article1 sourcesince Nov 4Last article 1h ago ·

Overview

AISummary of 1 article

Artificial Analysis says it now uses Time per Task as the headline metric for comparing model speed across roughly 1,000 serverless API endpoints.

The metric divides output tokens per Intelligence Index task by end-to-end output speed, so it measures how long a model takes to finish a task rather than how fast it streams tokens.

The source says default API effort settings cut Time per Task by about 3 to 5 times for GPT-6 Astra, GPT-6.1 Sol and Claude Opus 5.5, with lower Intelligence Index scores.

Written by AI from the articles below · updated Nov 4, 12:36 PM ET

Check the sources:

Latest developmentNov 4, 12:34 PM ET
Artificial Analysis makes Time per Task its headline API benchmark metric

Article timeline

The articles in this story. Times are ET.

Nov 4
  1. Artificial AnalysisOfficialPick
    Artificial Analysis makes Time per Task its headline API benchmark metric

    AIArtificial Analysis says it now uses Time per Task as the headline metric for comparing model speed on its roughly 1,000 serverless endpoints. The metric divides output tokens per Intelligence Index task by end-to-end output speed, which counts the full request from send to final token. The source says API default effort cuts Time per Task by about 3 to 5x for GPT-6 Astra, GPT-6.1 Sol and Claude Opus 5.5, with lower Intelligence Index scores.

    Image from @ArtificialAnlys's post

Heat Trend

Not enough continuous observations to show a trend yet.