Skip to the comparison
STET
Join the waitlist

GLM 5.3 Flash high vs Haiku 5.5 low

No clear quality winner: Haiku 5.5 low won 12, GLM 5.3 Flash high won 5, 7 tied; Haiku 5.5 low leads on cost and agent time: 1.6× cheaper and 2.7× faster.

24 tasksData from

Stet score

GLM 5.3 Flash highcodex · high reasoningRank 15 of 16

29.095% range 19.9–39.3

Haiku 5.5 lowclaude-code · low reasoningRank 13 of 16

38.995% range 32.6–45.49.9 points higher

A directional read: their 95% score ranges overlap.

Cost / task

GLM 5.3 Flash high$0.41
Haiku 5.5 low$0.251.6× cheaper

Agent time / task

GLM 5.3 Flash high26.7 min
Haiku 5.5 low9.9 min2.7× faster

Quality

Head-to-head matchups

GLM 5.3 Flash high won5

Tied7

Haiku 5.5 low won12

Haiku 5.5 low’s win share 67%, 95% range 51–81%Ties are within judge noise.

Task by task

GLM 5.3 Flash high solved 0 tasks that Haiku 5.5 low didn’t, and Haiku 5.5 low solved 2 that GLM 5.3 Flash high didn’t. Both solved 9; neither solved 13.

GLM 5.3 Flash high only 0

Haiku 5.5 low only 2

  • stet-commit-5d766ddeEquivalent to the merged fix
  • stet-43493c52Passed 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.

All 24 compared tasks
Task by task
TaskGLM 5.3 Flash highHaiku 5.5 lowSolved 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-66762b4cPassed testsEquivalent to the merged fixBoth
stet-15439c21Passed testsEquivalent to the merged fixBoth
stet-commit-5d766ddeFailedEquivalent to the merged fixHaiku 5.5 low only
stet-43493c52FailedPassed testsHaiku 5.5 low only
stet-commit-f0caded9Passed testsEquivalent to the merged fixBoth
stet-6f84e978FailedFailedNeither
stet-2450ca2dFailedFailedNeither
stet-commit-342d0ac0FailedFailedNeither
stet-commit-f8ae1177FailedFailedNeither
stet-commit-60a0e6c2FailedFailedNeither
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
MetricGLM 5.3 Flash highHaiku 5.5 low
Harnesscodexclaude-code
Reasoning efforthighlow
Rank15 of 1613 of 16
Stet score29.095% range 19.9–39.338.995% range 32.6–45.4
Solved9/249 passed, 0 judged equivalent11/247 passed, 4 judged equivalent
Cost / task$0.41$0.25
Agent time / task26.7 min9.9 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 16 setups: bmk_d4ba310fdf50db4ccb2092e3.

See all 16 setups on the leaderboardHow Stet measures