Openai / gpt-5.4

high

94.6% 193 / 204 fixtures · 1 run(s)
34,260 input / 37,710 total output / 33,625 reasoning within output tokens $0.65130000
93.1% 190 / 204 fixtures · 1 run(s)
40,545 input / 47,276 total output / 42,568 reasoning within output tokens $0.81050250
Loading reliability summary…
Pass Rate Delta
-1.5% Text: 94.6% → JSON: 93.1%
+2
Gained
JSON pass / text fail
−5
Lost
Text pass / JSON fail
188
Unchanged Pass
Both pass
9
Unchanged Fail
Both fail
Fixture Reliability Delta
Fixture Text JSON Delta
Benchmark Deltas
Benchmark Text JSON Delta
merge_conflicts 91.7% 83.3% -8.3%
rebase 91.7% 83.3% -8.3%
branch_cleanup 100% 91.7% -8.3%
reflog 100% 91.7% -8.3%
worktree_usage 66.7% 75% + 8.3%
blame_forensics 91.7% 91.7% + 0%
cherry_pick 83.3% 83.3% + 0%
commit_messages 100% 100% + 0%
commit_squash 100% 100% + 0%
git_bisect 100% 100% + 0%
git_clean 100% 100% + 0%
git_grep 100% 100% + 0%
git_log_format 100% 100% + 0%
git_show 100% 100% + 0%
stash_recovery 100% 100% + 0%
submodule_usage 91.7% 91.7% + 0%
tag_management 91.7% 91.7% + 0%
Changed Fixtures (7)