Frontier-level coding with strong CursorBench results, higher token efficiency, and a faster default variant.

Cursor

Cursor has released Composer 2, its latest AI coding model, now available in the Cursor editor. It achieves significant benchmark improvements over previous versions — scoring 61.3 on CursorBench, 61.7 on Terminal-Bench 2.0, and 73.7 on SWE-bench Multilingual, compared to 44.2, 47.9, and 65.9 for Composer 1.5. The model is priced at $0.50/M input and $2.50/M output tokens, with a faster variant at $1.50/M input and $7.50/M output. Improvements stem from a continued pretraining run and reinforcement learning on long-horizon coding tasks, enabling the model to handle tasks requiring hundreds of sequential actions.

Introducing Composer 2 · Cursor