Openai / gpt-oss-120b

medium

94.6% 193 / 204 fixtures · 1 run(s)
46,151 input / 38,354 total output / 30,140 reasoning within output tokens $0.01422596
90.2% 184 / 204 fixtures · 1 run(s)
46,601 input / 41,910 total output / 27,809 reasoning within output tokens $0.01380535
Loading reliability summary…
Pass Rate Delta
-4.4% Text: 94.6% → JSON: 90.2%
+6
Gained
JSON pass / text fail
−15
Lost
Text pass / JSON fail
178
Unchanged Pass
Both pass
5
Unchanged Fail
Both fail
Fixture Reliability Delta
Fixture Text JSON Delta
Benchmark Deltas
Benchmark Text JSON Delta
git_grep 100% 75% -25%
reflog 100% 75% -25%
tag_management 83.3% 91.7% + 8.3%
cherry_pick 83.3% 75% -8.3%
merge_conflicts 83.3% 75% -8.3%
stash_recovery 100% 91.7% -8.3%
submodule_usage 100% 91.7% -8.3%
blame_forensics 100% 100% + 0%
branch_cleanup 91.7% 91.7% + 0%
commit_messages 100% 100% + 0%
commit_squash 91.7% 91.7% + 0%
git_bisect 100% 100% + 0%
git_clean 100% 100% + 0%
git_log_format 100% 100% + 0%
git_show 100% 100% + 0%
rebase 83.3% 83.3% + 0%
worktree_usage 91.7% 91.7% + 0%
Changed Fixtures (21)