HyperCLOVA 82B

Closed weights NAVER,Search Solutions 82B parameters September 2021

No estimate

No hardware requirements for this model

The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation
NAVER,Search Solutions
Organisation type
Industry,Industry
Country
Korea (Republic of)
Published
10 September 2021
Authors
Boseop Kim, HyoungSeok Kim, Sang-Woo Lee, Gichang Lee, Donghyun Kwak, Dong Hyeon Jeon, Sunghyun Park, Sungju Kim, Seonhoon Kim, Dongpil Seo, Heungsub Lee, Minyoung Jeong, Sungjae Lee, Minsub Kim, Suk Hyun Ko, Seokhun Kim, Taeyong Park, Jinuk Kim, Soyoung Kang, Na-Hyeon Ryu, Kang Min Yoo, Minsuk Chang, Soobin Suh, Sookyo In, Jinseong Park, Kyungduk Kim, Hiun Kim, Jisu Jeong, Yong Goo Yeo, Donghoon …

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Language
Task
Language modeling/generation, Chat, Translation, Text classification
Approach
Self-supervised learning

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Parameters
82B

"We introduce a Korean in-context large-scale LM with 82B parameters, i.e., HyperCLOVA. This is the first discovery on near 100B-scale non-English LM." According to media reports, HyperCLOVA has 204B parameters (i.e. a different version than in the paper) https://m.koreaherald.com/view.php?ud=20210525000824

Training data
300,000,000,000 tokens

"However, experiments in Section 5.2 use the model trained with 300B tokens, as HyperCLOVA Studio provided the 39B and 82B models trained with 300B tokens." "We introduce HyperCLOVA, a large-scale Korean in-context learning-based LM with nearly 100B parameters, by constructing a large Korean-centric corpus of 560B tokens." Based on tokenizing the Hyperclova article itself using OpenAI's tiktoken BPE tokenizer (https://github.com/openai/tiktoken), there are 3285 tokens for 1069 words - about 3 …

Training compute

The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.

Training compute
1.5 × 10²³ FLOP

"For experiments in Section 4, the model trained with 150B is used for fair comparison, because not all models are finished training at the same iteration. However, experiments in Section 5.2 use the model trained with 300B tokens, as HyperCLOVA Studio provided the 39B and 82B models trained with 300B tokens." 82e9 connections * 2 FLOP/connection * 300e9 tokens * 3 backward pass = 1.476e23 FLOP Calculation using GPU time corroborates this: - "Our model is based on megatron-LM (Shoeybi et al., …

How it was established
Operation counting,Hardware

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Training hardware
NVIDIA A100
Chips used
1,024
Chip-hours
658,637
Wall-clock time
643 hours (26.8 days)

see compute notes

Hardware utilisation
MFU 20.0%

6ND method gives 1.476e23 FLOP needed to train the model. Actual use was 1,024 A100 GPUs for 13.4 days for 150B tokens; 300B model probably took twice as long. 26.8 * 24 * 3600 * 1024 * 3.12e14 = 7.3978e23 1.476e23 / 7.3978e23 MFU = 0.1995

Power draw
827.0 kW
Compute cost
$586,409

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights
Closed — provider access only
Model access
API access
Training code
Unreleased

"We introduce HyperCLOVA Studio, an interactive prompt engineering interface which provides GUI and API interfaces like the OpenAI playground1"

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Foundation model
Yes
Likely above 10²³ FLOP
Yes
Why it is tracked
SOTA improvement

"HyperCLOVA with our training configuration shows state-of-the-art in-context zero-shot and few-shot learning performances on various downstream tasks in Korean" I don't see any standard benchmark that they claim SOTA on

Record confidence
Confident
Citations
131

Sources

Where this record came from and when it was last checked.

Reference
What Changes Can Large-scale Language Models Bring? Intensive Study on HyperCLOVA: Billions-scale Korean Generative Pretrained Transformers
Last updated
25 May 2026

What the numbers mean

What this model is

HyperCLOVA 82B was published by NAVER,Search Solutions, in Korea (Republic of), in September 2021. The organisation is categorised as industry,Industry.

It works in Language, and is recorded as doing language modeling/generation, Chat, Translation, Text classification.

Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.

Training and provenance

The training run consumed about 1.5 × 10²³ FLOP, on NVIDIA A100. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

The training set ran to roughly 300,000,000,000 tokens.

Its inclusion criterion is sOTA improvement.

Answers

HyperCLOVA 82B — common questions

01

What GPU do I need to run HyperCLOVA 82B?

None. HyperCLOVA 82B is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.

02

Is HyperCLOVA 82B open source?

No. HyperCLOVA 82B has not had its weights published, so it exists only as a service controlled by its owner.

03

How many parameters does HyperCLOVA 82B have?

HyperCLOVA 82B has 82B parameters. "We introduce a Korean in-context large-scale LM with 82B parameters, i.e., HyperCLOVA. This is the first discovery on near 100B-scale non-English LM." According to media reports, HyperCLOVA has 204B parameters (i.e. a different version than in the paper) https://m.koreaherald.com/view.php?ud=20210525000824. That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.

04

Who created HyperCLOVA 82B?

HyperCLOVA 82B was published by NAVER,Search Solutions, based in Korea (Republic of), categorised as industry,Industry.

05

When was HyperCLOVA 82B released?

HyperCLOVA 82B was published in September 2021. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

06

What is HyperCLOVA 82B used for?

HyperCLOVA 82B works in Language, and is recorded as handling language modeling/generation, Chat, Translation, Text classification. These are the areas it was designed around; they describe intent rather than a hard boundary.

07

How much compute was used to train HyperCLOVA 82B?

Around 1.5 × 10²³ FLOP, on NVIDIA A100. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.

Source

Original publication

Record last updated 25 May 2026

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.