Skip to the comparison
STET
Run it on your repo

GPT 5.6 Luna low vs Opus 4.8 medium

GPT 5.6 Luna low leads on speed and cost; quality is too close to call.

Quality
+2.0 pts
Opus 4.8 medium
4–4 head to head, 15 tied
Speed
3.5× faster
GPT 5.6 Luna low
6.7 min vs 23.3 min per task
Cost
123.1× cheaper
GPT 5.6 Luna low
$0.05 vs $6.27 per task

24 tasksData from

Stet score

GPT 5.6 Luna lowcodex · low reasoningRank 15 of 17

45.095% range 39.1–50.4

Opus 4.8 mediummedium reasoningRank 12 of 17

47.095% range 39.0–55.32.0 points higher

A directional read: their 95% score ranges overlap.

Quality

Head-to-head matchups

GPT 5.6 Luna low won4

Tied15

Opus 4.8 medium won4

1 undecided
Opus 4.8 medium’s win share 52%, 95% range 42–64%Ties are within judge noise.

Task by task

GPT 5.6 Luna low solved 1 task that Opus 4.8 medium didn’t, and Opus 4.8 medium solved 3 that GPT 5.6 Luna low didn’t. Both solved 9; neither solved 10.

stet-commit-bbbbae09Solved by 15 of 17 setups

  • GPT 5.6 Luna lowFailed
  • Opus 4.8 mediumPassed 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
TaskGPT 5.6 Luna lowOpus 4.8 mediumSolved by
stet-232b50baPassed testsPassed testsBoth
stet-3169aefaPassed testsPassed testsBoth
stet-5276aa3fPassed testsPassed testsBoth
stet-89dfbc27Passed testsPassed testsBoth
stet-b28f4fe7Passed testsPassed testsBoth
stet-commit-bbbbae09FailedPassed testsOpus 4.8 medium only
stet-66762b4cEquivalent to the merged fixPassed testsBoth
stet-commit-5d766ddeEquivalent to the merged fixEquivalent to the merged fixBoth
stet-commit-f0caded9Equivalent to the merged fixEquivalent to the merged fixBoth
stet-43493c52Passed testsPassed testsBoth
stet-commit-f8ae1177FailedPassed testsOpus 4.8 medium only
stet-15439c21Equivalent to the merged fixFailedGPT 5.6 Luna low only
stet-2450ca2dFailedFailedNeither
stet-commit-342d0ac0FailedFailedNeither
stet-commit-60a0e6c2FailedPassed testsOpus 4.8 medium only
stet-commit-5e075641FailedFailedNeither
stet-commit-2826765eFailedFailedNeither
stet-commit-a8716b88FailedFailedNeither
stet-commit-e928166fFailedFailedNeither
stet-commit-30506f8eFailedFailedNeither
stet-commit-87814583FailedFailedNeither
stet-commit-d2ab0c25FailedFailedNeither
stet-commit-e41c2e2dFailedFailedNeither

Side by side

Side by side
MetricGPT 5.6 Luna lowOpus 4.8 medium
Harnesscodex—
Reasoning effortlowmedium
Rank15 of 1712 of 17
Stet score45.095% range 39.1–50.447.095% range 39.0–55.3
Solved10/246 passed, 4 judged equivalent12/2410 passed, 2 judged equivalent
Cost / task$0.05$6.27
Agent time / task6.7 min23.3 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 .
  • Both setups come from the same leaderboard run of 17 setups: bmk_d4ba310fdf50db4ccb2092e3.

See all 17 setups on the leaderboardHow Stet measures