Skip to content
Trending storyDeveloping

Artificial Analysis compares Gemini 4 Carbon and Claude Fable 5.2 on MKA tasks

6 articles1 sourcesince Nov 10Last article 1h ago ·

Overview

AISummary of 6 articles

Artificial Analysis reports that Gemini 4 Carbon (high) finished a building-plan audit deck task in 76 turns, about one hour, for roughly $8.

Claude Fable 5.2 (max) took 224 turns, about three hours, and roughly $68 on the same task. Fable 5.2's deck scored 87% against Carbon's 72%, and Carbon viewed images 118 times against Fable's 203.

In the same tests across ten MKA verticals, Artificial Analysis finds no single model dominates. Claude Fable 5.2 leads five verticals, including Software & Data, Engineering, Logistics & Mapping, Business, and Finance, where it shares the lead with Claude Sonnet 5.5. Gemini 4 Carbon (high) leads Healthcare.

Written by AI from the articles below · updated Nov 10, 1:28 PM ET

Check the sources:

Developments

3 developments

  1. Nov 10, 1:16 PM ET · 2 articles
    Artificial Analysis compares Gemini 4 Carbon and Claude Fable 5.2
    Artificial Analysis: Gemini 4 Carbon finishes a task in 76 turns for about $8 versus Claude Fable 5.2's 224 turns for $68
  2. Nov 10, 1:16 PM ET · 2 articles
    Artificial Analysis reports model leaders across MKA verticals
    Artificial Analysis: Claude Fable 5.2 leads five of ten MKA verticals in Artificial Analysis tests
  3. Nov 10, 1:16 PM ET · 2 articles
    Artificial Analysis reports Claude Fable 5.2's performance on an MKA task
    Artificial Analysis: Artificial Analysis reports Claude Fable 5.2 costs about $68 on one MKA task

Article timeline

The articles in this story. Times are ET.

Nov 10
  1. Artificial AnalysisOfficial
    Artificial Analysis finds Gemini 4 Carbon cheaper and faster than Claude Fable 5.2

    AIArtificial Analysis reports that Gemini 4 Carbon (high) completed a deck-generation task in 76 turns, about an hour, for roughly $8. Claude Fable 5.2 (max) took 224 turns, about three hours, and roughly $68 on the same task. Fable 5.2's deck scored higher, 87% against 72%, while Carbon viewed images 118 times against Fable's 203.

    Video from @ArtificialAnlys's post
  2. Artificial AnalysisOfficial
    Artificial Analysis reports Claude Fable 5.2 costs about $68 on one MKA task

    AIArtificial Analysis reports that Claude Fable 5.2 (max) needed 224 turns and 203 image views to finish one MKA task. The task audits building plans for door and window mismatches and reports the findings as a slide deck. The single task cost about $68.

    Video from @ArtificialAnlys's post
  3. Artificial AnalysisOfficial
    Claude Fable 5.2 leads five of ten MKA verticals in Artificial Analysis tests

    AIArtificial Analysis reports that model rankings vary widely across ten MKA verticals, with no single model dominating. Claude Fable 5.2 leads five verticals, including Software & Data, Engineering, Logistics & Mapping, Business, and Finance, where it shares the lead with Claude Sonnet 5.5 in Finance. Claude Opus 5.5 (max) leads Education and Construction, Gemini 4 Argon (high) leads Science, Gemini 4 Carbon (high) leads Healthcare, and Claude Sonnet 5.5 (max) narrowly leads Law & Government.

    Image from @ArtificialAnlys's post
  4. Artificial AnalysisOfficial
    Gemini 4 Carbon finishes a task in 76 turns for about $8 versus Claude Fable 5.2's 224 turns for $68

    AIArtificial Analysis reports that Gemini 4 Carbon (high) completed the same task in 76 turns, about one hour, and roughly $8. Claude Fable 5.2 (max) took 224 turns, about three hours, and roughly $68. Fable 5.2's deck scored higher, 87% versus 72%, and Carbon viewed images 118 times compared with Fable's 203.

    Video from @ArtificialAnlys's post
  5. Artificial AnalysisOfficial
    Claude Fable 5.2 leads five of ten MKA verticals in Artificial Analysis tests

    AIArtificial Analysis tests models across 10 MKA verticals and finds that no single model dominates. Claude Fable 5.2 leads five verticals, including Software & Data, Engineering, Logistics & Mapping, Business, and Finance, where Claude Sonnet 5.5 shares the lead in Finance. Claude Opus 5.5 (max) leads Education and Construction, Gemini 4 Argon (high) leads Science, Gemini 4 Carbon (high) leads Healthcare, and Claude Sonnet 5.5 (max) narrowly leads Law & Government.

    Image from @ArtificialAnlys's post
  6. Artificial AnalysisOfficial
    Artificial Analysis reports Claude Fable 5.2 costs about $68 on one MKA task

    AIArtificial Analysis reports that Claude Fable 5.2 (max) needed 224 turns and 203 image views to finish one MKA task. The task audits building plans for door and window mismatches and reports the findings as a slide deck. The single task cost about $68.

    Video from @ArtificialAnlys's post

Heat Trend

Not enough continuous observations to show a trend yet.