Sonnet 5.5 lands in Cursor, positioned near Opus-level

Cursor, the AI code editor, has added Anthropic's Sonnet 5.5 to the models its users can choose from, and its announcement calls it a strong model that performs on par with Opus in many tasks. Anthropic offers its Claude models in tiers, with Sonnet as the smaller model below Opus, so Cursor's claim is that the smaller model now keeps up with the larger one on many tasks.

That is Cursor's own assessment, and the post gives no numbers behind it. It does not say which tasks, how the comparison was made, or which Opus version it means, and by saying many tasks rather than all it leaves room for work where Opus stays ahead. The post includes an image whose contents were not available in the collected material, so any figures it may show are not reflected here. Measured comparisons come from elsewhere: Artificial Analysis's evaluation of Sonnet 5.5, posted earlier the same day, scored the model against Opus 5.5 on its own benchmark suite.