Megadose Models, research, and what moved.

Claude Fable 5.2 Scores 65 on Artificial Analysis, Highest So Far

· HN · Frontpage AI ·
The top benchmark result comes with a $26.54 average cost and 44.1 minutes per task.

Artificial Analysis evaluated Claude Fable 5.2 in its “Max, Default Fallback” configuration across nine evaluations. Its average task time compares with 12.9 minutes for models in a similar price tier. API pricing is $8 per million input tokens and $40 per million output tokens. The evaluation generated 275 million output tokens, against a comparable-model median of 72.8 million.

HN · Frontpage AI's note

score 8

Categories: Model Releases

Discussions

  • hn · 261 points · 184 comments
  • hn · 270 points · 190 comments
  • hn · 276 points · 199 comments