A.X K1

Closed weights SK Telecom 519B parameters December 2025

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
SK Telecom
Organisation type
Industry
Country
Korea (Republic of)
Published
30 December 2025

What it does

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

Domain
Language
Task
Code generation, Language modeling/generation, Translation, Instruction interpretation, Mathematical reasoning, Chat, Language modeling, Language generation, Text autocompletion
Numerical format
BF16

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
519B

MoE Architecture, Total parameters 519B, Active parameters 33B

Training data
tokens
Batch size
18,432

ramp up from 2,048, step size 1,024

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 H200
Chips used
1,536
Wall-clock time
1,536 hours (64 days)

GPU hours (1,536 GPUs)

Hardware utilisation
MFU 28.3% · HFU 92.7%
Compute cost
$9,642,947

How it is classified

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

Why it is tracked
Training cost
Record confidence
Confident

Sources

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

Reference
A.X K1
Last updated
31 December 2025

What the numbers mean

Background

A.X K1 was published by SK Telecom, in Korea (Republic of), in December 2025. industry is the category the publisher falls under.

It works in Language, and is recorded as doing code generation, Language modeling/generation, Translation, Instruction interpretation, Mathematical reasoning, Chat, Language modeling, Language generation, Text autocompletion.

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

Training and provenance

Its inclusion criterion is training cost.

Answers

A.X K1 — common questions

01

When was A.X K1 released?

A.X K1 was published in December 2025.

02

What is A.X K1 used for?

A.X K1 works in Language, and is recorded as handling code generation, Language modeling/generation, Translation, Instruction interpretation, Mathematical reasoning, Chat, Language modeling, Language generation, Text autocompletion. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

03

What GPU do I need to run A.X K1?

None. A.X K1 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.

04

Is A.X K1 open source?

The licensing for A.X K1 was never recorded in our source data. We treat unstated licensing as closed, because an unrecorded licence is not one to rely on.

05

How many parameters does A.X K1 have?

A.X K1 has 519B parameters. MoE Architecture, Total parameters 519B, Active parameters 33B. 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.

06

Who created A.X K1?

A.X K1 was published by SK Telecom, based in Korea (Republic of), categorised as industry.

Source

Original publication

Record last updated 31 December 2025

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.