Which has fewer bugs, Claude Opus 5 or GPT-5.6 Luna?

GPT-5.6 Luna had 19% fewer counted bugs per line than Claude Opus 5 (1.28 vs 1.58). GPT-5.6 Luna ranks #5 of 7 and Claude Opus 5 #6.

Which is used more, Claude Opus 5 or GPT-5.6 Luna?

Claude Opus 5, with 38.3% of the AI-written code reviewed by cubic in the week of Sep 14, 2026, against 1.7% for GPT-5.6 Luna.

Updated Sep 26, 2026

GPT-5.6 Luna had 19% fewer counted bugs per line than Claude Opus 5 (1.28 vs 1.58).

Claude Opus 5 and GPT-5.6 Luna compared on relative bug rate, by coding agent, bug type, and language, and on price and usage.
Bug rate
	Claude Opus 5	GPT-5.6 Luna
Relative bug rate	1.58	1.28
Rank	#6 of 7	#5 of 7
Details
Share of AI-written code, week of Sep 14	38.3%	1.7%

1.00 = average · lower is better

Which has fewer bugs, Claude Opus 5 or GPT-5.6 Luna?: GPT-5.6 Luna had 19% fewer counted bugs per line than Claude Opus 5 (1.28 vs 1.58). GPT-5.6 Luna ranks #5 of 7 and Claude Opus 5 #6.
Which is used more, Claude Opus 5 or GPT-5.6 Luna?: Claude Opus 5, with 38.3% of the AI-written code reviewed by cubic in the week of Sep 14, 2026, against 1.7% for GPT-5.6 Luna.