GPT 6 Luna low vs Opus 4.8 medium

No clear quality winner: Opus 4.8 medium won 14, GPT 6 Luna low won 5, 4 tied, 1 undecided; GPT 6 Luna low leads on cost and agent time: 132.2× cheaper and 3.4× faster.

24 tasksData from October 29, 2026

Stet score

GPT 6 Luna lowcodex · low reasoningRank 16 of 16

19.995% range 11.6–29.5

Opus 4.8 mediumclaude-code · medium reasoningRank 12 of 16

39.295% range 29.0–49.719.3 points higher

Other setups

Muse Wonder Full low

70.4

GPT 5.6 Luna max

70.3

SWE-2 Max high

64.3

GPT 6.1 Sol medium

62.5

Opus 5 medium

62.3

Grok 4.7 high

60.7

Pareto 26.9 medium

57.1

Haiku 5.5 max

55.5

GPT 6 Sol medium

55.4

GPT 6 Luna max

47.4

DeepSeek V4.1 Flash max

45.9

Haiku 5.5 low

38.9

Muse Spark 1.3 high

31.9

GLM 5.3 Flash high

29.0

A directional read: their 95% score ranges overlap.

Cost / task

GPT 6 Luna low$0.05132.2× cheaper

Opus 4.8 medium$7.05

Agent time / task

GPT 6 Luna low6.9 min3.4× faster

Opus 4.8 medium23.3 min

Quality

Head-to-head matchups

GPT 6 Luna low won5

Tied41 undecided

Opus 4.8 medium won14

Opus 4.8 medium’s win share 69%, 95% range 52–85%Ties are within judge noise.

Task by task

GPT 6 Luna low solved 0 tasks that Opus 4.8 medium didn’t, and Opus 4.8 medium solved 7 that GPT 6 Luna low didn’t. Both solved 5; neither solved 11.

GPT 6 Luna low only 0

Opus 4.8 medium only 7

stet-5276aa3fPassed tests
stet-commit-bbbbae09Passed tests
stet-66762b4cPassed tests
stet-commit-5d766ddeEquivalent to the merged fix
stet-43493c52Passed tests
stet-commit-f0caded9Equivalent to the merged fix
stet-6f84e978Passed tests

A task counts as solved when its tests passed or a judge found the fix equivalent to the merged change, as in the leaderboard’s Solved column. Stet score also weighs code quality, so it can order the two differently. 1 of the 24 tasks is left out because one of the two has no scored result on it.

All 23 compared tasks

Task by task
Task	GPT 6 Luna low	Opus 4.8 medium	Solved by
`stet-3169aefa`	Passed tests	Passed tests	Both
`stet-89dfbc27`	Passed tests	Passed tests	Both
`stet-b28f4fe7`	Passed tests	Passed tests	Both
`stet-232b50ba`	Passed tests	Passed tests	Both
`stet-5276aa3f`	Failed	Passed tests	Opus 4.8 medium only
`stet-commit-bbbbae09`	No change made	Passed tests	Opus 4.8 medium only
`stet-66762b4c`	Failed	Passed tests	Opus 4.8 medium only
`stet-15439c21`	Passed tests	Passed tests	Both
`stet-commit-5d766dde`	No change made	Equivalent to the merged fix	Opus 4.8 medium only
`stet-43493c52`	Failed	Passed tests	Opus 4.8 medium only
`stet-commit-f0caded9`	Failed	Equivalent to the merged fix	Opus 4.8 medium only
`stet-6f84e978`	Failed	Passed tests	Opus 4.8 medium only
`stet-2450ca2d`	Failed	Failed	Neither
`stet-commit-342d0ac0`	Failed	Failed	Neither
`stet-commit-f8ae1177`	No change made	Failed	Neither
`stet-commit-5e075641`	Failed	Failed	Neither
`stet-commit-2826765e`	Failed	Failed	Neither
`stet-commit-a8716b88`	Failed	Failed	Neither
`stet-commit-e928166f`	Failed	Failed	Neither
`stet-commit-30506f8e`	Failed	Failed	Neither
`stet-commit-87814583`	Failed	Failed	Neither
`stet-commit-d2ab0c25`	Failed	Failed	Neither
`stet-commit-e41c2e2d`	Failed	Failed	Neither

Side by side

Side by side
Metric	GPT 6 Luna low	Opus 4.8 medium
Harness	codex	claude-code
Reasoning effort	low	medium
Rank	16 of 16	12 of 16
Stet score	19.995% range 11.6–29.5	39.295% range 29.0–49.7
Solved	5/245 passed, 0 judged equivalent	13/2411 passed, 2 judged equivalent
Cost / task	$0.05	$7.05
Agent time / task	6.9 min	23.3 min

See which setup wins on your repo

Stet replays your merged PRs, scores quality above pass/fail, and delivers comparison reports through your agent.

Run it on your repo

About this data

24 tasks, each a merged change replayed from the repo as it was just before the merge.
Data from October 29, 2026.
Both setups come from the same leaderboard run of 16 setups: bmk_d4ba310fdf50db4ccb2092e3.

See all 16 setups on the leaderboard How Stet measures