OpenAI quietly updated its API pricing this week, and the developer internet did what it always does: lost its collective mind over fractions of a cent.
The company slashed rates on several models, with the flagship GPT-4o line getting the deepest cuts.
If you've been putting off building that AI-powered app because your cloud bill looked like a mortgage payment, this one's for you.
The headline number: GPT-4o input tokens dropped to $2.50 per million, down from $5.
Output tokens fell to $10 per million from $15.
The smaller GPT-4o mini is now comically cheap at 15 cents per million input tokens, which is roughly the cost of a gumball you'd find under a couch cushion in 1997.
For the non-developers in the room, "tokens" are how these models measure text.
A million tokens is roughly 750,000 words, or about three full novels.
So yeah, you could process the entire Harry Potter series through the mini model for less than the price of a sad airport sandwich.
Why does this matter to regular people who don't know what an API is?
Because every AI app you use, from your note-taking tool to that chatbot your bank added last month, runs on these prices.
When OpenAI cuts costs, the savings eventually trickle down to consumers in the form of cheaper subscriptions, free tiers that don't suck, and features that used to be "premium" suddenly appearing in the basic plan.
Anthropic, Google, and a growing pile of open-source models have been eating into OpenAI's lunch.
Meta's Llama models are free to run if you've got the hardware.
Google keeps undercutting on Gemini pricing.
This is a price war, and OpenAI is firing back with the only weapon that matters to businesses: your competitor's invoice.
Developers on Reddit and X wasted no time.
Some celebrated by immediately spinning up side projects they'd shelved.
Others did the math and realized their entire startup's unit economics just changed overnight.
A few cynics pointed out that cheaper tokens mean more aggressive rate limits and quieter model changes, which is the kind of paranoia that only comes from being burned before.
The practical takeaway for anyone building or buying AI products: costs are falling fast, and anything priced around last year's rates is probably overcharging you.
If you're paying for an AI tool that hasn't gotten cheaper or better in six months, it might be time to shop around. **Our take:** OpenAI slashing prices is great news for consumers and a brutal wake-up call for anyone whose business model was "we charge a lot because AI is expensive." The moat isn't the model anymore.
Final Thoughts
It's the product wrapped around it, and that's where most companies are about to get exposed.