EXAONE-4.5-33B: benchmark scores, price and ranking
EXAONE-4.5-33B from LGAI-EXAONE ranks #81 of 743 LLMs on the LLMs Tiger aggregate benchmark score (52.1), based on 2 benchmark results captured up to 2026-10-13.
- Organisation
- LGAI-EXAONE
- Type
- Open-weight / local model
- Open weights
- Yes
- Aggregate rank
- #81 of 743 (score 52.1)
Across the 2 benchmarks where at least ten models are scored, EXAONE-4.5-33B is in the top 10% on 0 and in the top half on 2.
EXAONE-4.5-33B benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| MMLU-Pro | 83.3% | #21 of 136 | HuggingFace MMLU-Pro leaderboard | 2026-10-13 |
| GPQA Diamond | 80.5% | #102 of 275 | HuggingFace GPQA Diamond leaderboard | 2026-10-13 |
About EXAONE-4.5-33B
EXAONE-4.5-33B is an open-weight model that can run on your own hardware from LGAI-EXAONE. No API price is listed for it yet. We hold 2 benchmark results for it across 1 area, and none is in the top tenth of its board.
Where EXAONE-4.5-33B stands
On our aggregate score, EXAONE-4.5-33B is #81 of 743 ranked models, in the top quarter. Its score is 52.1; the scale is centred on 50, so it sits 2.1 points above the typical model. That beats 89% of the ranked field. Among the 5 ranked LGAI-EXAONE models it is #1. The next one down is K-EXAONE-236B-A23B at 52.0.
EXAONE-4.5-33B, benchmark by benchmark
MMLU-Pro puts EXAONE-4.5-33B at 83.3%, #21 of 136. That beats 85% of the models tested, but 4.8 points short of Darwin-180B-RSI (88.1%). The results either side are Intern-S1 (83.5%) and Qwen3-235B-A22B-Instruct-2507 (83.0%). It covers 57-subject knowledge + reasoning. The board median is 55.3%, so EXAONE-4.5-33B is 28.0 points above it. Source: HuggingFace MMLU-Pro leaderboard, captured 2026-10-13.
On GPQA Diamond, EXAONE-4.5-33B reaches 80.5%. It ranks #102 of 275, ahead of 63% of the field, and 15.3 points behind GPT-6 Astra (max) at 95.8%. The results either side are MiniMax-M2.1 (80.8%) and LongCat-Flash-Thinking-2601 (80.5%). It covers expert-level science questions. The board median is 71.7%, so EXAONE-4.5-33B is 8.8 points above it. Source: HuggingFace GPQA Diamond leaderboard, captured 2026-10-13.
Strengths and weaknesses
Reasoning, maths and science: solid, ahead of 74% of the field on average across 2 benchmarks. Its best board here is MMLU-Pro (#21 of 136) and its weakest is GPQA Diamond (#102 of 275).
Reading these numbers
The 2 scores on this page come from 2 sources: HuggingFace GPQA Diamond leaderboard (1), HuggingFace MMLU-Pro leaderboard (1). No official leaderboard supplied any of them, so treat them as third-party measurements. With 2 results the aggregate rank can move a lot when one new score arrives.
Models ranked nearby
- #78 Claude Code + Opus 5 (52.1)
- #79 Qwen3.8-2.4T-A95B (52.1)
- #80 gpt-5.2-2025-12-11-high (52.1)
- #82 Claude Haiku 5.5 (max) (52.1)
- #83 NVIDIA-Nemotron-3-Super-120B-A12B-BF16 (52.0)
- #84 K-EXAONE-236B-A23B (52.0)
More from LGAI-EXAONE
- K-EXAONE-236B-A23B (#84, 52.0)
- EXAONE-3.5-32B-Instruct (#359, 48.4)
- EXAONE-3.5-7.8B-Instruct (#500, 46.6)
- EXAONE-3.5-2.4B-Instruct (#562, 45.7)
Compare it on the live leaderboard → · How the aggregate score works