Haiku 5.5 max leads on quality and agent time: 8.6 points higher and a little faster; GPT 6 Luna max leads on cost: 10.3× cheaper.
Stet score
GPT 6 Luna maxRank 8 of 14
50.095% range 40.1–59.3
Haiku 5.5 maxRank 6 of 14
58.695% range 47.5–69.08.6 points higher
GPT 6 Luna max solved 3 tasks that Haiku 5.5 max didn’t, and Haiku 5.5 max solved 3 that GPT 6 Luna max didn’t. Both solved 9; neither solved 8.
stet-232b50baPassed testsstet-66762b4cEquivalent to the merged fixstet-43493c52Passed testsstet-15439c21Passed testsstet-6f84e978Equivalent to the merged fixstet-2450ca2dPassed testsA 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.
| Task | GPT 6 Luna max | Haiku 5.5 max | 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 | Failed | GPT 6 Luna max only |
stet-5276aa3f | Passed tests | Passed tests | Both |
stet-commit-bbbbae09 | Passed tests | Passed tests | Both |
stet-15439c21 | Failed | Passed tests | Haiku 5.5 max only |
stet-66762b4c | Equivalent to the merged fix | Failed | GPT 6 Luna max only |
stet-commit-f0caded9 | Equivalent to the merged fix | Passed tests | Both |
stet-commit-5d766dde | Equivalent to the merged fix | Equivalent to the merged fix | Both |
stet-43493c52 | Passed tests | Failed | GPT 6 Luna max only |
stet-6f84e978 | Failed | Equivalent to the merged fix | Haiku 5.5 max only |
stet-commit-f8ae1177 | Passed tests | Passed tests | Both |
stet-2450ca2d | Failed | Passed tests | Haiku 5.5 max only |
stet-commit-342d0ac0 | Equivalent to the merged fix | Equivalent to the merged fix | Both |
stet-commit-5e075641 | Failed | Failed | Neither |
stet-commit-2826765e | Failed | Failed | Neither |
stet-commit-a8716b88 | 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 |
stet-commit-e928166f | Failed | Failed | Neither |
| Metric | GPT 6 Luna max | Haiku 5.5 max |
|---|---|---|
| Rank | 8 of 14 | 6 of 14 |
| Stet score | 50.095% range 40.1–59.3 | 58.695% range 47.5–69.0 |
| Solved | 12/248 passed, 4 judged equivalent | 12/249 passed, 3 judged equivalent |
| Passed review | 6/24 | 9/222 not recorded |
| Matched merged change | 11/24 | 10/222 not recorded |
| Cost / task | $0.37 | $3.85 |
| How cost was recorded | reported by its harness | reported by its harness on 23, estimated from tokens on 1 |
| Agent time / task | 59.7 min | 58.4 min |
| Ran in | Daytona sandbox | Daytona sandbox |
The leaderboard’s readings on groups of tasks. Each group holds fewer tasks than the overall standings, so treat these as leads to test on your repo.
11 tasks
11 tasks whose merged change was 2–149 lines in 1–2 files. The 7 small fixes below are among them.
| GPT 6 Luna max | Haiku 5.5 max | |
|---|---|---|
| Matched merged change | 7/11 | 7/101 not recorded |
| Passed review | 6/11 | 7/101 not recorded |
| Cost / task | $0.09 | $1.18 |
| Agent time / task | 19.0 min | 36.3 min |
7 tasks
7 tasks whose instructions start with “Fix”. Their merged changes were 4–29 lines in 1–2 files.
Both matched the merged change on 5 of 7.
Haiku 5.5 max passed review more often: 5 of 7 to GPT 6 Luna max’s 4.
| GPT 6 Luna max | Haiku 5.5 max | |
|---|---|---|
| Matched merged change | 5/7 | 5/7 |
| Passed review | 4/7 | 5/7 |
| Cost / task | $0.08 | $1.15 |
| Agent time / task | 17.1 min | 34.6 min |
13 tasks
13 tasks whose merged change ran over 150 lines or more than 3 files: 113–1,420 lines in 2–19 files.
| GPT 6 Luna max | Haiku 5.5 max | |
|---|---|---|
| Matched merged change | 4/13 | 3/121 not recorded |
| Passed review | 0/13 | 2/121 not recorded |
| Cost / task | $0.61 | $6.11 |
| Agent time / task | 94.2 min | 77.1 min |
Stet replays your merged PRs, scores quality above pass/fail, and delivers comparison reports through your agent.
Run it on your repocompare-n24-r3b.