Skip to the comparison

Opus 4.8 vs Opus 5

Opus 5 leads on quality, cost and agent time: 24.9 points higher, a little cheaper and a little faster.

24 tasksData from

Stet score

Opus 4.8Rank 11 of 14

40.295% range 30.3–50.3

Opus 5Rank 3 of 14

65.195% range 54.8–74.924.9 points higher

Their 95% score ranges don’t overlap.

Cost / task

Opus 4.8$7.05
Opus 5$6.78a little cheaper

Agent time / task

Opus 4.823.3 min
Opus 523.2 mina little faster

Task by task

Opus 4.8 solved 1 task that Opus 5 didn’t, and Opus 5 solved 2 that Opus 4.8 didn’t. Both solved 12; neither solved 9.

Opus 4.8 only 1

  • stet-commit-60a0e6c2Passed tests

Opus 5 only 2

  • stet-2450ca2dPassed tests
  • stet-commit-2826765eEquivalent to the merged fix

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
TaskOpus 4.8Opus 5Solved 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-15439c21Passed testsPassed testsBoth
stet-66762b4cPassed testsPassed testsBoth
stet-commit-f0caded9Equivalent to the merged fixPassed testsBoth
stet-commit-5d766ddeEquivalent to the merged fixEquivalent to the merged fixBoth
stet-43493c52Passed testsPassed testsBoth
stet-6f84e978Passed testsPassed testsBoth
stet-commit-f8ae1177FailedFailedNeither
stet-2450ca2dFailedPassed testsOpus 5 only
stet-commit-342d0ac0FailedFailedNeither
stet-commit-60a0e6c2Passed testsFailedOpus 4.8 only
stet-commit-5e075641FailedFailedNeither
stet-commit-2826765eFailedEquivalent to the merged fixOpus 5 only
stet-commit-a8716b88FailedFailedNeither
stet-commit-30506f8eFailedFailedNeither
stet-commit-87814583FailedFailedNeither
stet-commit-d2ab0c25FailedFailedNeither
stet-commit-e41c2e2dFailedFailedNeither
stet-commit-e928166fFailedFailedNeither

Side by side

Side by side
MetricOpus 4.8Opus 5
Rank11 of 143 of 14
Stet score40.295% range 30.3–50.365.195% range 54.8–74.9
Solved13/2411 passed, 2 judged equivalent14/2412 passed, 2 judged equivalent
Passed review8/231 not recorded9/231 not recorded
Matched merged change11/2413/24
Cost / task$7.05$6.78
How cost was recordedreported by its harnessreported by its harness
Agent time / task23.3 min23.2 min
Ran inlocal Dockersandbox not recorded

Where each setup is the better choice

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.

Changes up to 150 lines

11 tasks

11 tasks whose merged change was 2–149 lines in 1–2 files. The 7 small fixes below are among them.

Opus 5 matched the merged change more often: 10 of 11 to Opus 4.8’s 9.

Changes up to 150 lines
Opus 4.8Opus 5
Matched merged change9/1110/11
Passed review8/119/101 not recorded
Cost / task$1.41$1.50
Agent time / task12.8 min18.4 min

Small fixes

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 7 of 7.

Small fixes
Opus 4.8Opus 5
Matched merged change7/77/7
Passed review7/77/7
Cost / task$1.17$0.99
Agent time / task9.7 min15.9 min

Larger changes

13 tasks

13 tasks whose merged change ran over 150 lines or more than 3 files: 113–1,420 lines in 2–19 files.

Opus 5 matched the merged change more often: 3 of 13 to Opus 4.8’s 2.

Larger changes
Opus 4.8Opus 5
Matched merged change2/133/13
Passed review0/121 not recorded0/13
Cost / task$11.83$11.24
Agent time / task32.2 min27.2 min

From the leaderboard’s routing advice

  • Small fix: use Opus 5.

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 14 setups: compare-n24-r3b.