[whatmodel.sh](../../README.md) / HyperCLOVA X SEED 32B Think # HyperCLOVA X SEED 32B Think NAVER Cloud · released 2025-12 · 33.31B params · hyperclovax **score 62.0** ±1.8 · comparison only: no current run option verified · from 8 benchmark results across 2 sources · † no independently run results verified Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at ● marks HyperCLOVA X SEED 32B Think among every model on that task or test, worst to best. 
task         vs the others         rank  counts  tests
agent tasks  ▁▁▂▆█▄●▄▄▃▁▁   #154 of 354     85%      5
web search   ●▁▂▆▅█▅▄▁▂▁▁    #96 of 96*     11%      1
agent tasks · #154 of 354 · 85% of its score Beats [o3](o3.md), [Gemini 2.5 Pro](gemini-2-5-pro.md), [Gemini 3.1 Flash-Lite](gemini-3-1-flash-lite.md) here, though they rank higher overall. Trails [SmolLM3 3B](smollm3-3b.md), [Jamba 1.5 Large](jamba-1-5-large.md), [GPT-4o (2024-05-13)](gpt-4o-2024-05-13.md) here, though they rank lower overall. 
test                                result  vs the others in its run  difficulty            counts  why
Terminal-Bench Hard (maker)           9.9%  ▂▅▆▂▂●▂▂█▅▅▂ beats 13/27  very hard · avg 2.1%     24%  sharp · small run
τ²-bench Telecom (maker)             65.1%  ▁·▅█▆●▂▃▃▂▃▁ beats 52/89  very hard · avg 12%      20%  sharp · at its level · noisy
τ²-bench Retail (maker)              71.6%  ▁····▂▁▃▆●▃▁ beats 30/44  medium · avg 37%         16%  at its level
τ²-bench Airline (maker)             58.0%  ▁▁·▂▁▃▂▁●▅▂▃ beats 21/42  hard · avg 29%           13%  at its level
K-BrowseComp-Verified (Korean web…    2.3%  ●····█··▃█·▃ beats 0/8    very hard · avg 3.6%     11%  typical
web search · #96 of 96 · 11% of its score Trails [GPT-4.1 nano](gpt-4-1-nano-2025-04-14.md), [GPT-4o mini (2024-07-18)](gpt-4o-mini-2024-07-18.md), [Mistral Small 3.1 24B (2503)](mistral-small-3-1.md) here, though they rank lower overall. 
test                                result  vs the others in its run  difficulty            counts  why
K-BrowseComp-Verified (Korean web…    2.3%  ●····█··▃█·▃ beats 0/8    very hard · avg 3.6%     11%  typical
other tests (3) 
test               result  vs the others in its run    difficulty        counts  why
HellaSwag (maker)   57.2%  ▁▁▁▂▂▂●▆█▇▄▂ beats 36/135   medium · avg 63%    6.3%  at its level · small run
MMLU (maker)        87.7%  ▁▁▁▁▁▂▂▅██▆● beats 202/214  medium · avg 68%    5.8%  noisy · small run
PIQA (maker)        76.7%  ▁▁▁▂▂▃▅●▅▃▁▁ beats 59/123   easy · avg 83%      1.5%  blunt · small run
## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md) 
source                                    results  share of the score
HyperCLOVA X 32B Think (submitted / mak…        7   87% ▇▇▇▇▇▇▇▇▇▇
K-BrowseComp: A Web Browsing Agent Benc…        1   11% ▇
prior: pull toward the middle                        1% ▇
weighted average                                   = 62.0
Licences: HyperCLOVA X 32B Think ([CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)) · K-BrowseComp: A Web Browsing Agent Benchmark Grounded in Korean Contexts ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) ## How to run it No verified API or local download yet. Model repository and documentation: [huggingface.co/naver-hyperclovax](https://huggingface.co/naver-hyperclovax/HyperCLOVAX-SEED-Think-32B) [interactive version](https://whatmodel.sh/model/hyperclova-x-seed-32b-think) · [JSON](https://whatmodel.sh/api?format=json&all=1)