Korean Benchmark for Advanced Linguistic Tasks (KoBALT)
We show this table for reference; we do not rank on it.
Evaluates advanced Korean linguistic competence.
Benchmark score on KoBALT — November 13, 2026
We compile the KoBALT rows from provider self-reports. A.X K2 leads the table at 73%, followed by K-Exaone (61.8%). We do not use these results to rank models overall.
A.X K2
SK Telecom
Reported by SK TelecomK-Exaone
LG AI Research
Reported by LG AI Research2 modelsKoreanKorean-language benchmarkRefreshingDisplay onlyUpdated November 13, 2026
Benchmark score table (2 models)
All 2 results on this page were reported by the labs that built the models.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | A.X K2SK TelecomReported by SK Telecom | 73% | Reported by SK Telecom | Not reported | Open |
| 2 | K-ExaoneLG AI ResearchReported by LG AI Research | 61.8% | Reported by LG AI Research | Not reported | Closed |
- Lab-reported
- Benchmark-run
- Independent
About KoBALT
Tasks
Linguistics questions
Format
Advanced linguistics
Difficulty
Advanced linguistic phenomena
Freshness and provenance
Version
KoBALT
Refresh cadence
Static
Staleness state
Refreshing
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does KoBALT measure?
Evaluates advanced Korean linguistic competence.
Which model scores highest on KoBALT?
A.X K2 by SK Telecom currently leads with a score of 73% on KoBALT.
How many models are evaluated on KoBALT?
2 AI models have published results on KoBALT in the BenchLM catalog.
Compare top models on KoBALT
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.