← All benchmarks

Not tier-ranked

Composer 2.5

Cursor · In-IDE coding model, not general-purpose

Cursor's in-editor coding model — excellent at its one job, not built to be ranked against generalists.

3 of 10 benchmark runs completed

Why this tier

Composer 2.5 sits outside the main tier ranking because it isn't trying to compete as a general-purpose model — it's purpose-built for Cursor's editor, tuned specifically for in-context code edits and completions. The three runs we logged show exactly what you'd expect: it's very good at the narrow job it's designed for, and that's a different question from where it would land on a general reasoning tier list.

Strengths

  • Purpose-built for fast, accurate in-editor code edits
  • Tightly integrated with the Cursor workflow
  • Strong results on the coding-specific runs we logged

Worth knowing

  • Not a general-purpose model — comparing it directly to tiered models is apples-to-oranges
  • Only useful inside (or alongside) the Cursor workflow it's built for

Best for

  • Teams already using Cursor as their primary editor
  • Fast, in-context code edits rather than open-ended reasoning
  • A companion to a general-purpose model, not a replacement

Also in Not tier-ranked

Not sure which model fits?

We'll help you pick the right one for the job.

Model choice is rarely one-size-fits-all. Tell us what you're building and we'll tell you plainly what fits.