OpenAI’s latest move is one of those updates that feels simple at first, then starts making more sense the longer you think about it. The company is pushing Codex harder, and the new OpenAI’s ChatGPT Pro for Codex tier lands in a strange but very practical middle zone. It costs less than the old Pro plan, gives far more usage than Plus, and signals something OpenAI seems ready to acknowledge: many users don’t need the highest tier, but they definitely need more than the basics.
If you’ve ever hit a usage cap right when you were deep into real work, you already understand the appeal. That is the whole point here. No big AI spectacle. No dramatic model rollout. Just a smart adjustment to pricing and limits for people who rely on Codex often enough to care, but not enough to justify the premium plan.
Quick Highlights
- New ChatGPT Pro tier costs $100 a month in the US.
- It gives 5x more Codex usage than ChatGPT Plus.
- The older Pro plan still exists at the higher price.
- OpenAI is also rebalancing Plus limits for Codex.
- The change is meant for longer, high effort coding sessions.
Why OpenAI is doing this now
OpenAI isn’t hiding the reason. Codex has clearly caught on. Sam Altman said on X that the product is getting a lot of love, and the company is responding to demand by launching a $100 ChatGPT Pro tier. That’s a very OpenAI way of saying, “People want more access, and we’ve found a new price point that might actually work for them.”
There’s also a bigger business story underneath the announcement. Codex has reportedly reached 3 million active users, which is not a tiny side-project number. It’s enough to matter. OpenAI is also in a real race with Anthropic’s Claude Code, and when AI coding tools start pulling serious attention, pricing becomes part of the product war. It’s not just about which assistant writes code best. It’s about which one feels affordable, usable, and not annoyingly capped after a few heavy sessions.
That’s where this new tier starts to make sense. It’s not trying to be the cheapest thing around. It’s trying to be the “okay, this is the one I probably need” option.
What the new ChatGPT Pro tier actually gives you
The new ChatGPT Pro tier for Codex is priced at $100 per month in the US and Rs 10,699 in India. That’s still a serious subscription, no doubt. But it’s also almost half the price of the older Pro tier, which costs $200 in the US and Rs 19,900 in India. So
yes, OpenAI has created a cheaper premium tier without fully removing the expensive one.
And here’s the part that matters most: this new tier offers 5 times more Codex usage than ChatGPT Plus. OpenAI says it’s best for longer, high-effort Codex sessions. In plain language, that means if you’re using Codex for real work, not just a quick experiment or a tiny prompt here and there, this plan is aimed at you.
The company also says it includes the same broad perks as the pricier Pro plan, including unlimited access to Instant and Thinking models. So the reduction is not in the general ChatGPT experience. The big change is in Codex limits. That’s a subtle but important difference, because a lot of users don’t want to lose model access. They just want enough breathing room to code without constantly checking a counter like it’s a fuel gauge.
| Plan | Monthly price US | Monthly price India | Codex usage | Best for |
|---|---|---|---|---|
| ChatGPT Plus | $20 | Rs 1,999 | Baseline | Casual and lighter use |
| New ChatGPT Pro tier | $100 | Rs 10,699 | 5x Plus | Longer coding sessions |
| Older Pro plan | $200 | Rs 19,900 | 20x Plus | Heavy power users |
So what changed for Plus users?
This is where the update gets a little less cheerful. OpenAI has also changed the ChatGPT Plus limits for Codex. The company said the old promotion for existing Plus subscribers is ending, and as part of that, Codex usage in Plus is being rebalanced.
That sounds mild, but the practical meaning is probably what you’d expect: more sessions spread across a week, instead of long, deep daily sessions. OpenAI even said this shift should allow for more sessions over a week rather than extended usage on every single day. So the weekly pattern gets a little friendlier, but the daily ceiling may feel tighter.
In other words, Plus is still useful, but it’s less likely to satisfy someone who really leans on Codex as part of a work routine. If you’re coding occasionally, you may barely notice. If you’re in the middle of building, debugging, testing, and reworking things all day, you might very quickly bump into the new reality.
And honestly, that’s the trade-off OpenAI seems to be making. Give lighter users a plan that works. Keep the heavier users moving upward. It’s not particularly surprising, but it is a little bold in how directly it nudges people.
Why this matters for everyday users
For a beginner or intermediate user, all the talk about tiers and limits can sound a bit abstract. But if you use AI tools for learning to code, automating boring tasks, or building small projects, these changes affect the feel of the product immediately. The difference between “I can keep going” and “I’ve hit my cap again” is bigger than it sounds.
That’s why usage limits matter so much in AI tools. They aren’t just a spec. They shape how the software fits into your day. A generous limit turns AI into a real workflow companion. A tight limit turns it into a sometimes-helpful assistant that keeps stepping out of the room when you need it most.
The new tier seems designed for people who are stuck in that awkward middle zone: they don’t want to spend $200, but $20 isn’t really enough anymore either. That’s actually a pretty common spot for users to be in. And when a company finally acknowledges that middle, it usually means the product has grown up a bit.
OpenAI’s bigger strategy is hard to miss
There’s also a strategic angle here that’s worth noticing. OpenAI has been leaning harder into enterprise tools and Codex as it tries to improve profitability. At the same time, it recently shut down its video-generation platform Sora, which shows the company is willing to cut distractions if they don’t fit the main business focus. OpenAI reportedly wants to avoid “sidequests.”
That phrase is kind of funny, but the meaning is clear. The company wants fewer shiny detours and more serious products that can make money. Coding agents are a much cleaner business case than experimental features people might play with for a week and forget. If Codex keeps growing, pricing tiers like this become part of a long-term monetization strategy, not just a one-off experiment.
And the timing is interesting too. Anthropic’s revenue momentum is putting pressure on OpenAI, and competition has a way of making pricing more deliberate. When rivals are winning attention, the product that feels better balanced often ends up winning loyalty.
Should you care about the new Pro tier?
If you’re a light ChatGPT user, probably not that much. You might never need it. The Plus plan is still the more approachable entry point, and for many people it’ll be enough. But if you’ve been using Codex regularly and wishing it had more headroom, this new tier could be the sweet spot OpenAI is talking about.
Here’s the simple way to think about it:
- ChatGPT Plus is for casual or moderate use.
- New ChatGPT Pro tier is for users who need more Codex time without going all the way to the old Pro price.
- Older Pro still exists for people who need the most generous limits.
That’s not a bad lineup, actually. It’s a little more honest than the usual all-or-nothing subscription setup. The company is basically saying there’s a real audience in between hobbyist and power user, and that audience probably doesn’t want to pay twice as much just to keep working.
Still, there’s a tiny bit of caution worth keeping in mind. OpenAI is framing the change as a better fit for more users, but it’s also clearly steering people away from the lower-priced plan if they use Codex heavily. That’s not evil or unusual. It’s just pricing strategy. But it does mean the “cheaper” option isn’t necessarily the same value it used to be.
So, if you use Codex for serious projects, the question isn’t just whether the new tier is cheaper. It’s whether the extra budget actually buys you smoother work and fewer interruptions. For a lot of people, that’s probably the real test.
OpenAI’s new ChatGPT Pro tier for Codex feels like one of those updates that’s less about hype and more about fitting actual user behavior. And that’s refreshing, even if the company’s motives are obviously business-driven. Maybe the better question isn’t whether the new tier is perfect. It’s whether it finally gives users a more natural choice. If you’ve been running into Codex limits, this might be the first pricing move that actually feels designed around how people work, not just what a subscription chart wants to show.
