Skip to content
Trending storyDeveloping

Cursor adds Claude Fable 5.2 at $8/$40 per million tokens

2 articles1 sourcesince Nov 9Last article 6h ago ·

Overview

AISummary of 2 articles

Cursor says Claude Fable 5.2 is now available in its editor.

At max effort it scores 60% on CursorBench, the top result Cursor reports, and at default effort it scores 56.9%, up from 49.2% for Fable 5.1.

Cursor says Fable 5.2 costs $8 per million input tokens and $40 per million output tokens, 20% less than Fable 5.1. It also says effort levels behave differently from Fable 5.1, so users should start at the default rather than keep an old setting, and it links to an evals page comparing all effort levels.

Written by AI from the articles below · updated Nov 9, 9:00 PM ET

Check the sources:

Developments

2 developments

  1. Nov 9, 2:52 PM ET · 1 article
    Cursor reports Fable 5.2 pricing and effort-level differences
    Cursor: Cursor prices Fable 5.2 at $8 per million input tokens and $40 per million output tokens
  2. Nov 9, 2:52 PM ET · 1 article
    Claude Fable 5.2 becomes available in Cursor
    Cursor: Cursor adds Claude Fable 5.2, which scores 60% on CursorBench at max effort

Article timeline

The articles in this story. Times are ET.

Nov 9
  1. CursorOfficial
    Cursor prices Fable 5.2 at $8 per million input tokens and $40 per million output tokens

    AICursor says Fable 5.2 costs $8 per million input tokens and $40 per million output tokens, which is 20% less than Fable 5.1. It also says effort levels behave differently from Fable 5.1, so users should start at the default rather than carry over an old setting. Cursor links to an evals page for comparing every effort level.

  2. CursorOfficial
    Cursor adds Claude Fable 5.2, which scores 60% on CursorBench at max effort

    AICursor says Claude Fable 5.2 is now available in its editor. At max effort, Fable 5.2 is the top model on CursorBench at 60%. At default effort, it scores 56.9%, up from 49.2% for Fable 5.1.

    Image from @cursor_ai's post

Heat Trend

Not enough continuous observations to show a trend yet.