Are AI tokens the new signing bonus or just a cost of doing business?

This week, a topic that has been growing in Silicon Valley moved into the obvious: AI tokens as compensation. The idea is straightforward enough – rather than giving developers just salary, equity, and bonuses, companies will also give them a budget of AI tokens, units of which include powerful tools like Claude, ChatGPT, and Gemini. Use them to run agents, automate tasks, and run code. The point is that access to more computing makes engineers more productive, and that productive engineers are more important. It’s an investment in the person who owns them, it’s an idea.
Jensen Huang, the leather-jacketed CEO of Nvidia, seemed to capture everyone’s imagination when he presented the idea at the company’s annual GTC event earlier this week that developers should receive almost half of their base salary again – in tokens. His top people, by his calculations, could burn up to $250,000 a year on an AI computer. He called it a recruiting tool and predicted it would become the standard across Silicon Valley.
It’s not entirely clear where the idea was first, well, conceived. Tomasz Tunguz, a prominent VC in the Bay Area who runs Theory Ventures and specializes in 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 add up to “a quarter of engineering compensation.” Using data from the compensation-tracking website Levels.fyi, he put the average software engineer’s salary at $375,000. Add $100,000 in tokens and you’re up to $475,000 fully loaded – meaning roughly one in five dollars is now accounted for.
That is not a coincidence. Agent AI has taken off, and the release of OpenClaw in late January accelerated the conversation considerably. OpenClaw is an open source AI assistant designed to work continuously – it checks tasks, generates mini-agents, and works through to-do lists while its user sleeps. It’s part of a broader shift toward “agent” AI, meaning systems that don’t just respond to information but take sequential actions automatically over time.
The practical effect is that the use of tokens has exploded. Where someone writing an essay might spend 10,000 tokens in an afternoon, a developer using a swarm of agents can blow millions a day – automatically, in the background, without typing a word.
This weekend, the New York Times put together an insightful look at the so-called tokenmaxxing trend, finding that developers at companies including Meta and OpenAI compete on internal leaderboards that track token usage. Open token budgets are quietly becoming common practice, the paper reports, the way dental insurance or a free lunch once was. One Ericsson engineer in Stockholm told the Times that he probably spends more on Claude than he earns in salary, although his employer picks up the tab.
Perhaps tokens will indeed become the fourth pillar of engineering compensation. But developers may want to hold the line before accepting this as an outright win. More tokens may mean more power in the short term, but given how fast things are evolving, it doesn’t necessarily mean more job security. First, a large token allocation comes with high expectations. If a company successfully funds the cost of a second engineer’s computer on your behalf, the obvious pressure is to produce at double the cost (or more).
Techcrunch event
San Francisco, CA
|
October 13-15, 2026
And there’s a muddy problem underneath that: when the company’s token spending per employee approaches or exceeds that employee’s salary, the accounting financial logic begins to look different from its financial team. When a computer does a job, the question of how many people need to coordinate it becomes difficult to avoid.
Jamaal Glenn, an East Coast MBA from Stanford and a former VC turned CFO of financial services, similarly points out that what may be seen as a bonus may be a smart way for companies to increase the perceived value of the compensation package without raising cash or equity – things that actually involve the employee in the long run. Your token budget is not valid. It doesn’t matter. It does not appear in your next offer negotiations how the basic salary or allowance works. If companies successfully create tokens as payment, they may find it easier to keep cash comp flat while pointing to the growing computing capital as proof of investment in their people.
That’s a good thing for the company. Whether it’s a benefit to developers depends on questions many developers don’t yet have enough knowledge to answer.



