The token doesn't need a Chinese name, but the business behind it does.

Author: Curry, Deep Tide TechFlow

You may have noticed something recently: people have started discussing what to call tokens.

Professor Yang Bin of Tsinghua University published an article titled "Determining the Chinese translation of Token is imminent"; related translation questions on Zhihu have received 250,000 views, and the comments section is full of ideas.

For the past two or three years, the domestic AI community has been using the term "Token" without any apparent problem. Why is a Chinese name suddenly needed?

The direct reason may be that after the Spring Festival this year, ordinary people learned for the first time that tokens cost money.

OpenClaw has transformed AI from chatting into doing work, burning through hundreds of thousands of tokens for a single task, causing bills to skyrocket; various cloud vendors have also announced price increases, with tokens now serving as the unit of measurement.

At the same time, tokens began appearing in places where they shouldn't have been before.

At the GTC conference, Nvidia President Jensen Huang said that some people in Silicon Valley are already asking "how many tokens will I get for this job" during interviews, and he suggested that tokens be included in engineers' salaries.

OpenAI founder Sam Altman went even further, believing that tokens will replace basic income for all citizens, and that what everyone receives will not be money, but computing power.

Data from the National Bureau of Data Science shows that China's daily token consumption rose from 100 billion in early 2024 to over 40 trillion in September 2025, and reached 180 trillion in February of this year. The People's Daily published an article at the beginning of the year titled "A Casual Discussion of Tokens," explaining the meaning of this term to readers.

Once a technical term is included in cloud service bills, job payroll packages, and official statistics, it can no longer be referred to as English.

The question is, what should it be called?

If this were merely a translation issue, the answer would have been available long ago. In 2021, the Chinese academic community gave "Token" a name: lexical element.

But nobody cared, because at that time, "Token" was just an internal term in the tech world.

Things are different now.

The word "token" is inherently a versatile term. People in the cryptocurrency world called it "cryptocurrency token," those in security called it "token," and those in AI called it "word element." The same English word, depending on which direction the Chinese translation leans towards, becomes its domain.

Thus began a battle to name the token.

Business requires a voice

How to translate a word is usually the job of linguists. But among those involved in naming this time, there were almost no linguists.

The name with the most buzz right now is "Zhiyuan".

The most enthusiastic promoter is an AI media outlet called "New Zhiyuan". If the Chinese name for Token is "Zhiyuan", the company's brand name will overlap with basic industry terminology, which means that every article discussing Token is giving it free advertising.

Their own promotional article ended quite candidly: "We suggest translating Token as the new industry consensus: Zhiyuan, leaving the word 'new' to us."

According to the same article, Wang Xiaochuan, founder of Baichuan Intelligence, commented: "The name Zhiyuan is quite good."

He's working on large-scale models, so calling the token "Zhiyuan" is a good idea. Each computation of the model no longer produces a billing unit, but rather a "basic unit of intelligence."

Selling tokens is selling traffic, while selling Zhiyuan is selling intelligence; their valuation stories are completely different.

Professor Yang Bin of Tsinghua University proposed the concept of "module element," where "module" corresponds to a model. Whoever owns the large model controls the production rights of the "module element." By using a name that leans towards models, the pricing power shifts to model companies.

Some people advocate calling it a "symbol token." Going back to the most basic definition in computer science, a token is a symbol processing unit that has nothing to do with intelligence or models.

Technically the cleanest, but the proposer was an independent technical author, without company backing or capital support, and thus had almost no voice in this discussion.

The direction a name leans towards determines the direction the industry narrative takes, and the direction money flows in.

A distant example is the day Facebook changed its name to Meta; the "metaverse" transformed from a science fiction concept into a company's valuation story. A more recent example is that China consumes 180 trillion tokens daily, the highest in the world, but what this term is called, how it's defined, and who should define it remain unresolved...

The world's largest consumer of tokens hasn't even decided what to call the things they consume.

However, this word already has a Chinese name.

In 2021, Professor Qiu Xipeng of the School of Computer Science at Fudan University translated "Token" as "词元" (cíyuán), which was accepted by the academic community and included in textbooks. At that time, no one discussed this because tokens were worthless.

Tokens are valuable now.

It's the unit of account for cloud services, the source of revenue for large-scale AI companies, and a core indicator for national statistics on the scale of the AI ​​industry. So the media came, the bigwigs came, the professors came, each bringing their preferred names and the rationale behind them.

Translation has never been the problem. The problem is when this word started to become valuable.

Jensen Huang did not participate in the Chinese naming discussion at GTC. He did something simpler: he held up a championship belt that read "Token King" and announced that the data center was the Token Factory.

Who produces the token, who defines the token. What the name is, he doesn't care.

Token, Land Acquisition and Minting

Therefore, the real point worth considering is not which translation is better.

After the term "calorie" was established, the entire food industry's pricing, labeling, and regulatory systems were built around it. After the definition of "data traffic" was established in China's telecommunications industry, operators billed based on data traffic, competed based on data traffic, and designed packages based on data traffic. The entire business model revolved around these two words for more than a decade.

Tokens are now following the same path.

It has become the billing unit for cloud services, a revenue metric for large-scale AI companies, and a core indicator for measuring the scale of the AI ​​industry at the national level. The VC community has even begun discussing whether investment payments could be made directly in tokens.

Once a word becomes a unit of measurement for money, naming it is no longer a translation, but a minting of coins.

Call it "Intelligent Yuan," and the power to mint currency belongs to the AI ​​narrative; whoever tells the story of intelligence benefits. Call it "Model Yuan," and the power to mint currency belongs to the model companies; whoever has the big models can print money. Call it "Character Yuan," and the power to mint currency returns to the technology itself, but the technology itself will not speak for itself.

The academic definition of "lexicon" in 2021 was ignored, not because of poor translation, but because this "coin" was not valuable at that time.

Now it's valuable, and everyone wants to have their name engraved on it.