ChatGPT Just Got a $100 Middle Tier — And Vibe Coders Are the Reason

There’s a new ChatGPT subscription sitting between the $20 Plus plan and the $200 Pro plan. And honestly, it’s about time.

OpenAI announced a new $100-a-month Pro tier built specifically for people who need more coding capacity without paying the full premium price. If you’ve been hitting your usage limits mid-project and staring at a reset countdown, this new plan is aimed directly at you.

Vibe Coding Broke the Plus Plan

Here’s the thing about AI coding tools. Writing code eats through tokens — the basic unit of AI generation — much faster than regular chatting does.

When you ask ChatGPT to plan and build an entire app feature, you’re not just having a conversation. You’re generating massive amounts of structured output. That burns through your monthly limits at a completely different rate than asking for recipe ideas.

So plenty of users on the $20 Plus plan were running out of capacity well before the month ended. Their only options were waiting for a reset or jumping all the way to $200. That’s a pretty brutal gap.

What the New $100 Plan Actually Gives You

The new tier sits squarely in the middle, and the math is pretty clean.

You get Codex usage limits five times higher than the Plus plan. Plus, you get access to everything included in the existing $200 Pro plan — including the full ChatGPT Pro model and unlimited access to Instant and Thinking models. The only real difference from the pricier tier is usage volume. The $200 plan offers 20 times the Plus limits, compared to five times for this new option.

For anyone who codes regularly but doesn’t need industrial-scale output, that distinction matters a lot.

Codex Usage Exploded This Year

OpenAI’s timing here isn’t random. The company reported that Codex usage has grown more than 70% month-over-month in 2025. That’s a staggering jump.

Vibe coding — the practice of describing what you want an AI to build and letting it handle the actual code — went from a niche curiosity to a mainstream workflow practically overnight. Anthropic’s Claude Code grabbed viral attention at the end of 2025. OpenAI’s Codex followed with its own boom in early 2026.

More users coding means more tokens consumed. More tokens consumed means Plus plan limits get hit faster. OpenAI’s response is essentially to add a lane on the highway instead of raising speed limits for everyone.

Anthropic Already Plays This Game

Worth noting: OpenAI didn’t invent this pricing structure. Anthropic’s Claude subscriptions already follow an identical pattern.

Claude Max 5x costs $100 a month and offers five times the capacity of the base $20 plan. Claude Max 20x runs $200 and delivers 20 times the capacity. The new ChatGPT tier mirrors that structure almost exactly.

That’s probably not a coincidence. Both companies are clearly seeing the same user behavior — heavy coders hitting limits and needing a smarter middle option.

Is It Worth It?

If you’re a casual ChatGPT user, probably not. The $20 Plus plan handles general tasks, writing, and light AI assistance without any real strain.

But if you’re using Codex to build projects, prototype ideas, or automate workflows, the math shifts quickly. Hitting your limits mid-project is genuinely frustrating. And paying $200 for limits you’ll never fully use doesn’t make much sense either.

The $100 tier fills a real gap. It’s not glamorous pricing, but it solves an actual problem that a growing number of people have. For serious vibe coders who live inside Codex, this is probably the plan that finally makes sense.