benchgap
Calibration

CursorBench 4.0 → CursorBench 3.2

CursorBench 3.2 is estimated from CursorBench 4.0 with a Michaelis–Menten curve fitted on 9 models measured on both: y = 1.1637·x / (0.30024 + x), R² = 0.87, cross-validated error 2.2 pp. It is used for 2 estimates.

Estimated modelCursorBench 4.0CursorBench 3.2Source
Grok 4.746.3%70.6%estimated ± 2.2 pp, high confidence
Muse Spark 1.341.6%67.6%estimated ± 2.2 pp, high confidence