Skip to the comparison
STET
Get a free trial

GPT 6 Luna max vs Haiku 5.5 low+guidance

GPT 6 Luna max leads on quality: 8.4 points higher; Haiku 5.5 low+guidance leads on cost and agent time: 1.3× cheaper and 4.8× faster.

24 tasksData from

Stet score

GPT 6 Luna maxcodex · max reasoningRank 8 of 15

50.895% range 41.2–60.08.4 points higher

Haiku 5.5 low+guidanceclaude-code · low reasoningRank 10 of 15

42.495% range 33.2–51.7

A directional read: their 95% score ranges overlap.

Cost / task

GPT 6 Luna max$0.37
Haiku 5.5 low+guidance$0.291.3× cheaper

Agent time / task

GPT 6 Luna max59.7 min
Haiku 5.5 low+guidance12.4 min4.8× faster

Quality

Head-to-head matchups

GPT 6 Luna max won11

Tied8

Haiku 5.5 low+guidance won5

GPT 6 Luna max’s win share 62%, 95% range 46–77%Ties are within judge noise.

Task by task

GPT 6 Luna max solved 3 tasks that Haiku 5.5 low+guidance didn’t, and Haiku 5.5 low+guidance solved 0 that GPT 6 Luna max didn’t. Both solved 9; neither solved 12.

GPT 6 Luna max only 3

  • stet-commit-5d766ddeEquivalent to the merged fix
  • stet-commit-f8ae1177Passed tests
  • stet-commit-342d0ac0Equivalent to the merged fix

Haiku 5.5 low+guidance only 0

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.

All 24 compared tasks
Task by task
TaskGPT 6 Luna maxHaiku 5.5 low+guidanceSolved by
stet-3169aefaPassed testsPassed testsBoth
stet-89dfbc27Passed testsPassed testsBoth
stet-b28f4fe7Passed testsPassed testsBoth
stet-232b50baPassed testsPassed testsBoth
stet-5276aa3fPassed testsPassed testsBoth
stet-commit-bbbbae09Passed testsPassed testsBoth
stet-66762b4cEquivalent to the merged fixPassed testsBoth
stet-15439c21FailedFailedNeither
stet-commit-f0caded9Equivalent to the merged fixEquivalent to the merged fixBoth
stet-commit-5d766ddeEquivalent to the merged fixFailedGPT 6 Luna max only
stet-43493c52Passed testsPassed testsBoth
stet-6f84e978FailedFailedNeither
stet-commit-f8ae1177Passed testsFailedGPT 6 Luna max only
stet-2450ca2dFailedFailedNeither
stet-commit-342d0ac0Equivalent to the merged fixFailedGPT 6 Luna max only
stet-commit-60a0e6c2FailedFailedNeither
stet-commit-5e075641FailedFailedNeither
stet-commit-2826765eFailedFailedNeither
stet-commit-a8716b88FailedFailedNeither
stet-commit-30506f8eFailedFailedNeither
stet-commit-87814583FailedFailedNeither
stet-commit-d2ab0c25FailedFailedNeither
stet-commit-e41c2e2dFailedFailedNeither
stet-commit-e928166fFailedFailedNeither

Side by side

Side by side
MetricGPT 6 Luna maxHaiku 5.5 low+guidance
Harnesscodexclaude-code
Reasoning effortmaxlow
Rank8 of 1510 of 15
Stet score50.895% range 41.2–60.042.495% range 33.2–51.7
Solved12/248 passed, 4 judged equivalent9/248 passed, 1 judged equivalent
Cost / task$0.37$0.29
Agent time / task59.7 min12.4 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 15 setups: bmk_d4ba310fdf50db4ccb2092e3.

See all 15 setups on the leaderboardHow Stet measures