Apple Cuts Employee AI Token Budgets as It Pushes Siri AI

Is the bill really that high?

The cut stands out because Apple is the most valuable listed technology company. Its accounts show R&D spending of $45.765 billion in its last fiscal year, up 32%, a rise Apple attributes partly to AI investment. Employees say few people used their full $300 daily allowance, so actual savings may be small. Others fear the limit is a prelude to deeper cuts or layoffs, as at other tech firms.

Nvidia takes the opposite view. Jensen Huang said on the All-In podcast that he would be worried if an engineer paid $500,000 spent less than $250,000 on tokens. Asked whether Nvidia spends around $2 billion a year on tokens for engineers, he replied: "We're trying." Nvidia sells the chips that run those models, so more tokens means more sales.

Tokenmaxxing hits the brakes

Apple is not alone. Meta saw an internal race to burn the most tokens, and Amazon shut down a similar ranking in May; its later memo discourages tokenmaxxing, though the company says it set no limits. At Microsoft, one employee spent $28,000 in tokens in a single month, and vice president Jay Parikh wrote internally that tokenmaxxing "is not what we are optimizing".

Revolut is taking another route. About 2,000 of its engineers now code by default with GLM, an open-weight Chinese model that can run on the company's own servers. Infrastructure chief David Abitbol said Revolut pays a tenth of what frontier models cost, and the bill is fixed rather than rising with each request. Previously, the $300 monthly per engineer for Claude Code did not last the month.