K-EXAONE (Reasoning) — benchmark results
Provider: LG AI. Released 2026-01-12. Access: Open.
Unified ELO 1640 ± 32, rank #408 of 1841 rated models, from 55 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA AIME 2025 | 90.33 | Accuracy (%) | 90.7 |
| AA LiveCodeBench | 76.83 | Pass@1 (%) | 88.3 |
| AA MMLU-Pro | 83.75 | Accuracy (%) | 87.2 |
| AA IFBench | 64.69 | Accuracy (%) | 76.3 |
| AA-LCR | 55.7 | Score (self-reported) | 71.8 |
| AA Humanity's Last Exam | 13.07 | Accuracy (%) | 69.8 |
| AA CritPt | 1.12 | Accuracy (%) | 69.7 |
| AA GPQA Diamond | 78.28 | Accuracy (%) | 69 |
| AA TAU-2 Bench | 74.27 | Accuracy (%) | 66 |
| AA Long Context Reasoning | 55.67 | Accuracy (%) | 65.1 |
| Artificial Analysis Intelligence Index | 22.1 | Intelligence Index | 63.6 |
| Tau3 Banking | 14.23 | Success Rate (%) | 63.5 |
Interactive version: theaggregate.ai/model?slug=k-exaone-reasoning · How It Works · Data refreshed daily, snapshot 2026-07-25.