KoBALT-700 (syntax)
KoreanHigher is better
The syntax part of KoBALT-700, a set of Korean linguistics questions written by experts: multiple-choice items on agreement, argument structure, embedded clauses, ellipsis, and word-order changes. Horangi runs a fixed 100-question subset rather than the full dataset, so its numbers are not directly comparable with full-dataset scores. Scores run from 0 to 100%. Higher is better.
Top score90.0%GPT-5.6 Sol
Models tested48
Last updated2026.08.03
Model release date (newest on the right)
Best score so farHigher on the chart is better
RankModelSettingReleasedPrice/1MScore
- 1max-effort2026.07.09$1690.0%
- 2high-effort2025.12.17$2.3888.0%

- 2xhigh-effort2026.04.23$2488.0%
- 2enable-thinking2026.04.02$0.3488.0%

- 5high-effort2026.06.09$4087.0%
- 52026.05.19$7.1387.0%

- 5high-effort2026.02.19$9.5087.0%

- 8max-effort2026.07.09$9.5085.0%
- 9max-effort2026.07.09$0.9584.0%
- 9high-effort2026.05.27$2084.0%
The "harness" is the agent program the AI used to carry out the task. The same model can score very differently depending on its harness and reasoning effort.