$39.6 billion—breaking Wall Street records.

Many people think that the biggest winners of the AI era will be model companies like OpenAI and Anthropic.
But in reality, the truly most profitable are still those quant “machines” equipped with mathematics, computing power, systems engineering, and capital efficiency.
And Jane Street, at its core, is a super AI agent in the financial world.
A company with more than 3,000 people, it is one of the world’s most important quantitative market makers—active across 45 countries and more than 200 trading platforms—spanning multiple asset classes such as ETFs, stocks, options, bonds, and forex. (p2 shows Jane Street’s office scene—think about how you fight daily against this group of profit-making agents.)
In 2025, Jane Street paid employees $9.38 billion in compensation. Based on approximately 3,500 employees, average compensation per person is about $2.68 million, which is roughly 18.29 million RMB.
Jane Street’s per-capita compensation is about 7 times that of Goldman Sachs. Even for new hires, the quant trader positions at Jane Street for fresh graduates have a base salary of around $300,000, not including bonuses.
After DeepSeek became popular, Jane Street is often discussed alongside Liang Wenfeng and Fantasia Quant.
Because the two have too-similar underlying temperaments.
Both are low-key.
Both started out in quant.
Both believe in mathematics, models, engineering, and computing power.
Neither follows traditional internet narratives.
Behind them are groups of people who are accustomed to modeling, calculating, and placing bets amid uncertainty.
Liang Wenfeng ultimately incubated DeepSeek. Jane Street didn’t step into building big models itself—it chose to make a heavy bet on the AI industry chain.
It invested in Anthropic, and it also appears behind AI companies such as CoreWeave and Thinking Machines Lab.