Benchmark

DeepSway

An AI benchmark focused on software engineering tasks. Models are required to complete bug fixes, feature implementations, or refactoring within a given codebase, and automated tests determine whether the task is passed.

Timeline (last 90 days)

Oct 13

Meta claims MuseWonder Full scores 56 on the Artificial Analysis Intelligence Index, only two points behind Claude Opus 5.5, and scores 79 on the DeepSway software engineering benchmark.

Unverified50%

Sep 10

Muse Spark 1.3, released by Meta, achieved a high score of 75.4 on the DeepSway code leaderboard.

Unverified50%

All Facts (2)

Unverified

Meta claims MuseWonder Full scores 56 on the Artificial Analysis Intelligence Index, only two points behind Claude Opus 5.5, and scores 79 on the DeepSway software engineering benchmark.

50%Unverified

Muse Spark 1.3, released by Meta, achieved a high score of 75.4 on the DeepSway code leaderboard.

50%

Source Articles

Meta MuseWonder Hands-On: Benchmark Scores Near Opus 5.5, But Real-World Coding Falls ShortOct 13