# K-EXAONE 2.0 Benchmark Scores & Performance

> K-EXAONE 2.0 by LG AI Research has 27 source-displayable benchmark rows but no public overall score. It remains unranked.

Canonical page: https://benchlm.ai/models/k-exaone-2-0

Last updated: September 10, 2026

## Model Details

| Property | Value |
|----------|-------|
| Creator | LG AI Research |
| Source Type | Open Weight |
| Reasoning Type | Reasoning |
| Context Window | 262K |
| Official model card | [LG AI Research K-EXAONE 2.0 model card](https://huggingface.co/LGAI-EXAONE/K-EXAONE-2.0-750B-A37B) |
| Overall Score | Coming soon |
| Overall Rank | Unranked |

## Family & Coverage

- Family: K-EXAONE 2.0
- Variant: base
- Benchmarks covered: 27 of 434
- Related earlier model: [K-Exaone](/models/k-exaone)
- Coverage note: BenchLM currently has partial benchmark coverage for this model, so the overall score is conservative.

## Agentic Benchmarks

| Benchmark | Score |
|-----------|-------|
| [Terminal-Bench 2.1](/benchmarks/terminalbench21) | 43.8% |
| [Claw-Eval](/benchmarks/claw-eval) | 77.7% |
| [GDPval-AA](/benchmarks/gdpvalaanormalized) | 20.9% |

## Coding Benchmarks

| Benchmark | Score |
|-----------|-------|
| [SciCode](/benchmarks/scicode) | 37.4% |
| [SWE-bench Verified](/benchmarks/swe-bench-verified) | 68.2% |
| [Terminal-Bench 2.1](/benchmarks/terminalbench21) | 43.8% |
| [AA-SciCode](/benchmarks/aascicode) | 42.0% |

## Reasoning Benchmarks

| Benchmark | Score |
|-----------|-------|
| [AA-LCR](/benchmarks/lcr) | 56.2% |
| [CritPt](/benchmarks/critpt) | 0.9% |

## Knowledge Benchmarks

| Benchmark | Score |
|-----------|-------|
| [MMLU-Pro](/benchmarks/mmlu-pro) | 83.5% |
| [GPQA-D](/benchmarks/gpqa-diamond) | 82.2% |
| [HLE](/benchmarks/hle) | 18.3% |
| [MMMLU](/benchmarks/mmmlu) | 86.6% |
| [Artificial Analysis Intelligence Index](/benchmarks/artificialanalysis) | 19.7% |
| [AA-GPQA Diamond](/benchmarks/aagpqadiamond) | 82.9% |
| [AA-HLE](/benchmarks/aahle) | 18.6% |
| [AA-Omniscience Index](/benchmarks/aaomniscienceindex) | -6.6% |
| [AA-Omniscience Accuracy](/benchmarks/omniscienceaccuracy) | 13.1% |
| [AA-Omniscience Hallucination Rate](/benchmarks/omnisciencehallucinationrate) | 22.6% |

## Instruction Following Benchmarks

| Benchmark | Score |
|-----------|-------|
| [IFEval](/benchmarks/ifeval) | 92.4% |
| [IFBench](/benchmarks/ifbench) | 72.6% |

## Multilingual Benchmarks

| Benchmark | Score |
|-----------|-------|
| [PolyMath](/benchmarks/polymath) | 71.3% |

## Korean Benchmarks Benchmarks

| Benchmark | Score |
|-----------|-------|
| [KMMLU-Pro](/benchmarks/kmmlupro) | 69.1% |
| [CLIcK](/benchmarks/click) | 84.2% |

## Mathematics Benchmarks

| Benchmark | Score |
|-----------|-------|
| [AIME26](/benchmarks/aime2026) | 92.3% |
| [HMMT Feb 2026](/benchmarks/hmmtfeb2026) | 78.4% |
| [IMOAnswerBench](/benchmarks/imoanswerbench) | 78.6% |

## Other LG AI Research Models

- [K-Exaone](/models/k-exaone) - Score: 42.82
- [Exaone 4.0 32B](/models/exaone-4-0-32b) - Score: 38.5
- [Exaone 4.0 1.2B](/models/exaone-4-0-1-2b) - Score: 37.91