This week, a topic that has been making the rounds in Silicon Valley came to the fore: AI tokens as compensation. The idea is pretty simple: Instead of giving engineers just salary, equity and bonuses, companies would also give them a budget of AI tokens, the computational units that power tools like Claude, ChatGPT and Gemini. Spend them to run agents, automate tasks, and execute code. The argument is that access to more computing makes engineers more productive, and that more productive engineers are worth more. The idea is an investment in the person who owns them.
Jensen Huang, Nvidia’s leather-jacket-wearing CEO, seemed to capture everyone’s imagination when he floated the idea at the company’s annual GTC event earlier this week that engineers should receive about half their base salary again, in tokens. Your best people, by your calculations, could burn out $250,000 a year in AI computing. He called it a recruiting tool and predicted it would become a standard across Silicon Valley.
It’s not entirely clear where the idea first emerged. Tomasz Tunguz, a renowned Bay Area venture capitalist who runs Theory Ventures and focuses on AI, data and SaaS startups, and whose writing on all things data has gained a loyal following over the years, was talking about this in mid-February, writing that tech startups were already adding inference costs like “fourth component to engineering compensation.” Using data from compensation tracking site Levels.fyi, he estimated the salary of a software engineer in the top quartile at $375,000. Add $100,000 in tokens and you have $475,000 fully loaded, meaning it’s now estimated to be roughly one dollar in five.
That’s not a coincidence. Agent AI has been taking off and the OpenClaw release at the end of January he considerably accelerated the conversation. OpenClaw is an open source AI assistant designed to run continuously: performing tasks, spawning subagents, and working on a to-do list while the user sleeps. It’s part of a broader shift toward “agent” AI, that is, systems that not only respond to prompts but autonomously take sequences of actions over time.
The practical consequence is that token consumption has skyrocketed. While someone writing an essay can spend 10,000 tokens in an afternoon, an engineer running a swarm of agents can spend millions in a day, automatically, in the background, without typing a word.
This weekend, the New York Times had produced a smart look in the so-called tokenmaxxing trend, finding that engineers at companies like Meta and OpenAI are competing on internal leaderboards that track token consumption. Generous token budgets are quietly becoming a standard job benefit, the newspaper reported, like dental insurance or a free lunch once were. An Ericsson engineer in Stockholm told the Times that he probably spends more on Claude than he earns in salary, although his employer foots the bill.
Perhaps tokens really will become the fourth pillar of engineering compensation. But engineers may want to hold their ground before accepting this as an easy victory. More tokens may mean more power in the short term, but given how quickly things are evolving, it doesn’t necessarily mean more job security. On the one hand, a large token allocation comes with high expectations. If a company effectively funds a second engineer’s computing on its behalf, the implicit pressure is to produce at twice the speed (or more).
Technology event
San Francisco, CA
|
October 13-15, 2026
And there’s a more confusing problem beneath that: At the point where a company’s token spending per employee approaches or exceeds that employee’s salary, the financial logic of the count starts to look different to its finance team. If computing is doing the work, the question of how many humans need to coordinate becomes harder to avoid.
Jamaal Glenn, East Coast-based Stanford MBA and former VC turned financial services CFO, similarly points out That what may seem like a benefit may be a clever way for companies to inflate the apparent value of a compensation package without increasing cash or equity, things that actually increase for an employee over time. Its symbolic budget is not consolidated. It is not appreciated. It doesn’t appear in your next offer negotiation like a base salary or capital grant does. If companies successfully normalize tokens as payment, they may find it easier to keep cash compensation stable while also pointing to a growing IT allocation as proof of investment in their people.
It’s good business for the company. Whether it is a good deal for the engineer depends on questions that most engineers do not yet have enough information to answer.
