CPM-2

Closed weights Tsinghua University,Beijing Academy of Artificial Intelligence / BAAI 11B parameters June 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
Tsinghua University,Beijing Academy of Artificial Intelligence / BAAI
Organisation type
Academia,Academia
Country
China
Published
24 June 2021
Authors
Zhengyan Zhang, Yuxian Gu, Xu Han, Shengqi Chen, Chaojun Xiao, Zhenbo Sun, Yuan Yao, Fanchao Qi, Jian Guan, Pei Ke, Yanzheng Cai, Guoyang Zeng, Zhixing Tan, Zhiyuan Liu, Minlie Huang, Wentao Han, Yang Liu, Xiaoyan Zhu, Maosong Sun

What it does

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

Domain
Language
Task
Language 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
11B

"we pretrain two models: an encoder-decoder bilingual model with 11 billion parameters (CPM2) and its corresponding MoE version with 198 billion parameters. "

Training data
69,615,000,000 tokens

"We pre-train our model on WuDaoCorpus (Yuan et al., 2021), which contains 2.3TB cleaned Chinese data as well as 300GB cleaned English data." 2300GB * 167M tokens/GB + 300GB * 267M tokens/GB = 464 billion tokens https://docs.google.com/document/d/1G3vvQkn4x_W71MKg0GmHVtzfd9m0y3_Ofcoew0v902Q/edit#heading=h.ieihc08p8dn0

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
3.6 × 10²² FLOP

epoch count not stated C = 6 FLOP * 11B * 464B =

How it was established
Operation counting

How it is classified

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

Record confidence
Likely
Citations
101

Sources

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

Reference
CPM-2: Large-scale Cost-effective Pre-trained Language Models
Last updated
25 May 2026

What the numbers mean

Where it came from

CPM-2 was published by Tsinghua University,Beijing Academy of Artificial Intelligence / BAAI, in the country recorded as China, during June 2021. The category the publisher falls under is academia,Academia.

It works in the domain of Language, and is recorded as performing the task of language generation.

Its weights were never published, so it can only be reached through its provider. No graphics card changes that.

What went into building it

Training it took a computation budget of roughly 3.6 × 10²² FLOP. That figure measures what producing the model cost, and has no bearing on how fast it answers.

Training consumed a corpus of around 69,615,000,000 tokens of text.

Answers

CPM-2 — common questions

01

CPM-2— when was it released?

It was published in June 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.

02

CPM-2— what is it used for?

It works in the domain of Language, and is recorded as handling the task of language generation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

03

CPM-2— how much compute was used to train it?

Training consumed around 3.6 × 10²² 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.

04

CPM-2— what GPU do I need to run it?

None. This 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.

05

CPM-2— is it open source?

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

06

CPM-2— how many parameters does it have?

It has a parameter count of 11B. "we pretrain two models: an encoder-decoder bilingual model with 11 billion parameters (CPM2) and its corresponding MoE version with 198 billion parameters. ". 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.

07

CPM-2— who created it?

It was published by Tsinghua University,Beijing Academy of Artificial Intelligence / BAAI, based in China, an organisation categorised as academia,Academia.

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.