LongNet

Closed weights Microsoft,Xi’an Jiaotong University 2.7B parameters July 2023

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
Microsoft,Xi’an Jiaotong University
Organisation type
Industry,Academia
Country
United States of America, China
Published
5 July 2023
Authors
Jiayu Ding, Shuming Ma, Li Dong, Xingxing Zhang, Shaohan Huang, Wenhui Wang, Nanning Zheng, Furu Wei

What it does

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

Domain
Language
Task
Language modeling/generation

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
2.7B

2.7B

Training data
300,000,000,000 tokens

2.7B model uses 300B tokens from The Stack, others only use 40B.

Epochs
1

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
4.9 × 10²¹ FLOP

2.7B params * 300B tokens * 6 = 4.86e21 Note: not sure if there are very long sequences in the training data that would affect this calculation. Per paper, complexity of their attention mechanism scales linearly with sequence length.

How it was established
Operation counting

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
Unreleased
Training code
Open source

MIT license https://github.com/microsoft/unilm/tree/master/longnet

How it is classified

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

Record confidence
Confident
Citations
242

Sources

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

Reference
LongNet: Scaling Transformers to 1,000,000,000 Tokens
Last updated
25 May 2026

What the numbers mean

What this model is

LongNet was published by Microsoft,Xi’an Jiaotong University, in United States of America, in July 2023. industry,Academia is the category the publisher falls under.

It works in Language, and is recorded as doing language modeling/generation.

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

How it was trained

Training it took roughly 4.9 × 10²¹ FLOP of computation — a measure of what producing the model cost, not of how fast it answers.

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

Answers

LongNet — common questions

01

Who created LongNet?

LongNet was published by Microsoft,Xi’an Jiaotong University, based in United States of America, categorised as industry,Academia.

02

When was LongNet released?

LongNet was published in July 2023. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

03

What is LongNet used for?

LongNet works in Language, and is recorded as handling language modeling/generation. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

04

How much compute was used to train LongNet?

Around 4.9 × 10²¹ FLOP. 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.

05

What GPU do I need to run LongNet?

None. LongNet 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.

06

Is LongNet open source?

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

07

How many parameters does LongNet have?

LongNet has 2.7B parameters. 2.7B. 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.

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.