DeepSeek V4.1 Flash leads on quality: 6.5 points higher; Haiku 5.5 low leads on cost and agent time: 2.0× cheaper and 5.0× faster.
Stet score
DeepSeek V4.1 FlashRank 9 of 14
47.895% range 37.5–58.06.5 points higher
Haiku 5.5 lowRank 10 of 14
41.395% range 34.8–47.8
DeepSeek V4.1 Flash solved 1 task that Haiku 5.5 low didn’t, and Haiku 5.5 low solved 0 that DeepSeek V4.1 Flash didn’t. Both solved 11; neither solved 12.
stet-commit-342d0ac0Equivalent to the merged fixA 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.
| Task | DeepSeek V4.1 Flash | Haiku 5.5 low | 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 | Passed tests | Passed tests | Both |
stet-commit-bbbbae09 | Passed tests | Passed tests | Both |
stet-15439c21 | Passed tests | Equivalent to the merged fix | Both |
stet-66762b4c | Passed tests | Equivalent to the merged fix | Both |
stet-commit-f0caded9 | Passed tests | Equivalent to the merged fix | Both |
stet-commit-5d766dde | Passed tests | Equivalent to the merged fix | Both |
stet-43493c52 | Passed tests | Passed tests | Both |
stet-6f84e978 | Failed | Failed | Neither |
stet-commit-f8ae1177 | Failed | Failed | Neither |
stet-2450ca2d | Failed | Failed | Neither |
stet-commit-342d0ac0 | Equivalent to the merged fix | Failed | DeepSeek V4.1 Flash only |
stet-commit-60a0e6c2 | Failed | Failed | Neither |
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 | DeepSeek V4.1 Flash | Haiku 5.5 low |
|---|---|---|
| Rank | 9 of 14 | 10 of 14 |
| Stet score | 47.895% range 37.5–58.0 | 41.395% range 34.8–47.8 |
| Solved | 12/2411 passed, 1 judged equivalent | 11/247 passed, 4 judged equivalent |
| Passed review | 8/24 | 8/24 |
| Matched merged change | 10/231 not recorded | 9/24 |
| Cost / task | $0.49from 23/24 tasks | $0.25 |
| How cost was recorded | estimated from tokens on 23, not recorded on 1 | reported by its harness |
| Agent time / task | 49.3 min | 9.9 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.
Both matched the merged change on 7 of 11.
DeepSeek V4.1 Flash passed review more often: 8 of 11 to Haiku 5.5 low’s 7.
| DeepSeek V4.1 Flash | Haiku 5.5 low | |
|---|---|---|
| Matched merged change | 7/11 | 7/11 |
| Passed review | 8/11 | 7/11 |
| Cost / task | $0.21 | $0.02 |
| Agent time / task | 34.9 min | 2.8 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.
DeepSeek V4.1 Flash passed review more often: 6 of 7 to Haiku 5.5 low’s 5.
| DeepSeek V4.1 Flash | Haiku 5.5 low | |
|---|---|---|
| Matched merged change | 5/7 | 5/7 |
| Passed review | 6/7 | 5/7 |
| Cost / task | $0.19 | $0.01 |
| Agent time / task | 34.6 min | 2.7 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.
| DeepSeek V4.1 Flash | Haiku 5.5 low | |
|---|---|---|
| Matched merged change | 3/121 not recorded | 2/13 |
| Passed review | 0/13 | 1/13 |
| Cost / task | $0.751 not recorded | $0.45 |
| Agent time / task | 61.5 min | 15.9 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.