Name: GPT-5.6 Luna Pro
Author: OpenAI

GPT-5.6 Luna Pro

OpenAI · 2026-07-09

GPT-5.6 Luna Pro is the higher-compute Pro variant of GPT-5.6 Luna, represented as a distinct Atlas model so benchmark results reported under the Pro identity are not collapsed into standard Luna. OpenAI exposes Pro through the gpt-5.6-luna API model with reasoning.mode set to pro rather than through a separate API model ID.

Benchmark scores

Benchmark	Score
AHK-Eval	94.4
AHK-Eval Algorithms	100.0
AHK-Eval Data Structures	100.0
AHK-Eval Date and Time	83.3
AHK-Eval Easy	100.0
AHK-Eval Hard	91.7
AHK-Eval Hidden Cases	96.7
AHK-Eval Mid	91.7
AHK-Eval Numbers	100.0
AHK-Eval Parse Success	100.0
AHK-Eval Regex	83.3
AHK-Eval Strings	100.0
BuseyBench SVG	5.7
CadQueryEval	76.0
CadQueryEval — Bounding Box	92.0
CadQueryEval — Chamfer Distance	92.0
CadQueryEval — Code Execution	92.0
CadQueryEval — Component Count	92.0
CadQueryEval — Hausdorff 95p	88.0
CadQueryEval — STL Creation	92.0
CadQueryEval — Volume	88.0
CadQueryEval — Watertight Mesh	84.0
Capability	161.3
ChessBench	24.0
ChessBench — Fork	3.3
ChessBench — Hanging Piece	10.0
ChessBench — Mate in 1	86.7
ChessBench — Mate in 2	16.7
ChessBench — Pin	3.3
Eval Connections	95.0