Reading up on CursorBench

Cursor improves its coding agent by iterating on context management, evaluation metrics, and model-specific harness customization, treating the harness as a software product.

Cursor uses real-time reinforcement learning on live user interactions to ship improved Composer model checkpoints every five hours.

Cursor released Composer 2, a frontier-level coding model achieving strong benchmarks (CursorBench 61.3, Terminal-Bench 61.7) at competitive pricing.

Cursor trains Composer to generate its own compact summaries mid-task, reducing context errors by 50% while using 80% fewer tokens than traditional prompting.

Cursor uses a hybrid online-offline eval system, CursorBench, built from real developer sessions to better distinguish model quality than public benchmarks.