SWE-Bench Multimodal Benchmark Leaderboard | Phaseo

SWE-Bench Multimodal

AI benchmark results and model performance

Type: percentageCode

View benchmark source

Recorded Results

4

Average Score

45.25%

Scores Over Time

Individual benchmark scores plotted by date.

Models Using This Benchmark

Organisation	Model	Reported	Top Score	Info	Self Reported	Source
Anthropic	Claude Sonnet 5	30 Jun 2026	28.10%	SWE-bench Multimodal; internal harness; average over five trials	Yes	Source
Anthropic	Claude Opus 4.7	16 Apr 2026	34.50%	-	Yes	Source
Anthropic	Claude Mythos Preview	07 Apr 2026	59%	Multimodal (internal implementation)	Yes	Source
Anthropic	Claude Opus 5	24 Jul 2026	59.40%	Anthropic internal harness; adaptive thinking at max effort; average of 5 trials	Yes	Source