Apple Cuts Some Engineers' Daily AI Budgets to $200

Apple has cut how much some of its engineers can spend on AI models each day, from $300 to $200, Bloomberg reported on November 4. The cut came in October.

Paying for Claude or ChatGPT seats for your own team? You’ve probably had this same talk.

Apple declined to comment, The Next Web reports.

What Apple is changing

More cuts are planned, according to people familiar with the change:

• Some engineering teams could drop again, to $100 a day.
• Many staff outside engineering already have a $100 daily cap.
• Apple has considered cutting that to $25 for some teams.
• Anyone who needs more can ask senior leaders for an exemption.

The budgets cover tokens, the units of text a model reads and writes. Coding burns through a lot of them, MacRumors explains, so engineers get the biggest allowance.

Apple staff can use Claude, ChatGPT, Gemini and other models.

But some teams are now told to switch to cheaper models or on-device AI, the kind that runs on the device itself, so you don’t pay per token.

Will Apple save much? Staff told Bloomberg few people used their full $300.

Apple isn’t the only one

TechCrunch calls this year’s rush “tokenmaxxing”. Companies pushed staff to use AI as much as possible, and then the token bills came due.

Sound familiar? You’re not alone.

The Next Web lists others that now cap AI use:

• Microsoft set AI budgets per division in July.
• Tesla caps staff at $200 a week.
• Uber limits each AI coding tool to $1,500 a month.
• Accenture caps Claude at $29 a month.

Meanwhile, Apple’s research and development costs rose 32% in fiscal 2026, on higher AI and infrastructure costs.

Some Apple staff worry the cuts point to layoffs, after more than 200 job cuts in August.

CEO John Ternus, in charge since September 1, has pushed for a leaner Apple.

Why it matters

You probably don’t spend $200 a day on AI.

Still, your problem has the same shape at a five-person agency or a solo SaaS: coding agents eat tokens fast, and you don’t notice until the invoice lands.

So do what Apple did, at your own scale:

• Set a monthly cap per person or per tool, and write it down.
• Use a cheaper model for routine jobs, like GPT-6.1 Terra at half the price of Sol, and keep your top model for the hard problems.
• Check which seats nobody fully uses.

My read: the cap matters less than the switch to cheaper models. Most of your team’s AI work doesn’t need the priciest model, whether it’s a meta description or a support reply.

That said, I wouldn’t cut so hard that your team stops using AI.

What does your team spend a day? If you can’t answer that, start there.

Common Questions

How much can Apple engineers spend on AI each day?

Up to $200 a day for some engineers since October 2026, down from $300, Bloomberg reported. Some engineering teams could drop to $100, and Apple has considered $25 for some teams outside engineering.

Which AI models can Apple employees use at work?

Anthropic’s Claude, OpenAI’s ChatGPT, Google’s Gemini and other models, according to MacRumors and The Next Web. Some Apple teams are now told to use cheaper or on-device models instead.

Which other companies cap employee AI spending?

Microsoft set division-level AI budgets in July 2026, Tesla caps staff at $200 a week, Uber limits each AI coding tool to $1,500 a month and Accenture caps Claude at $29 a month, The Next Web reports.