EXAONE 4.5 33B — benchmark results
LG AI Research's open 33B vision-language model with hybrid attention (research license). Provider: LG AI. Released 2026-04-09. Access: Open.
Unified ELO 1609 ± 28, rank #461 of 1776 rated models, from 35 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| K-MetBench | 75.9 | Accuracy (self-reported) | 81 |
| AA GPQA Diamond | 79.39 | Accuracy (%) | 72.3 |
| AA IFBench | 57.96 | Accuracy (%) | 69.8 |
| AA TAU-2 Bench | 78.07 | Accuracy (%) | 68.4 |
| AA Humanity's Last Exam | 11.63 | Accuracy (%) | 66.8 |
| Artificial Analysis Intelligence Index | 22.96 | Intelligence Index | 65.1 |
| CritPt | 0.3 | Accuracy (self-reported) | 65 |
| AA-LCR | 49.3 | Score (self-reported) | 63.3 |
| AA Omniscience - Science, Engineering & Mathematics | 28.8 | Accuracy (%) | 60.4 |
| AA Terminal-Bench Hard | 20.45 | Accuracy (%) | 59.9 |
| AA CritPt | 0.29 | Accuracy (%) | 58.7 |
| AA Long Context Reasoning | 49.33 | Accuracy (%) | 57.7 |
Interactive version: theaggregate.ai/model?slug=exaone-4-5-33b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.