GPT-5 is Here, But OpenAI’s Limits Spark a Subscriber Revolt
The promise of artificial intelligence just got a significant upgrade – and a hefty dose of controversy. OpenAI’s GPT-5 has arrived, boasting substantial leaps in reasoning, coding, and tool utilization. But unlike previous releases, access isn’t uniform. While GPT-5 is available to all ChatGPT users, including those on the free tier, the tiered usage limits are already fueling backlash, particularly from paying subscribers who find themselves with less access than before.
A Two-Tiered Intelligence System
OpenAI’s decision to gate access based on subscription level is a strategic move to manage demand for its most powerful model. Free users receive a limited number of high-intelligence responses – a mere 10 messages every 5 hours, supplemented by a single “GPT-5 Thinking” message per day. ChatGPT Plus subscribers fare better, with up to 80 messages every 3 hours, but face automatic switching to a less capable “GPT-5 mini” once the limit is reached. Team and ChatGPT for Business plans enjoy unlimited access, highlighting the premium placed on consistent, high-level AI capabilities.
Understanding the Usage Limits
Here’s a breakdown of the current GPT-5 usage limits on ChatGPT:
- Free Tier: 10 messages every 5 hours + 1 GPT-5 Thinking message/day. Switches to GPT-5 mini after limit.
- ChatGPT Plus: 80 messages every 3 hours. Switches to GPT-5 mini after limit.
- ChatGPT Plus/Team: 200 GPT-5 Thinking messages/week (manual selection).
- ChatGPT for Business: Unlimited access to GPT-5 models.
Crucially, responses generated while the model automatically switches to GPT-5 Thinking don’t count against the primary limits. However, this small consolation hasn’t appeased many.
The Plus Subscriber Backlash: A Step Backwards?
The core of the discontent lies with ChatGPT Plus subscribers. Many are voicing frustration over the reduced access to “Thinking” models compared to previous iterations. Previously, Plus users enjoyed unlimited access to models like o3 and o4-mini, allowing for more complex and nuanced interactions. Now, they’re capped at 200 weekly messages for GPT-5 Thinking. The sentiment is palpable on social media:
“ChatGPT literally got worse for every single Plus user today. There’s no way to reliably get thinking models anymore. Before we had o4-mini, o4-mini-high and o3. Now we have GPT-5 Thinking with 200 messages per week and a router that exclusively routes you to some small and…” – @scaling01
Another user succinctly summarized the feeling: “We went from having solid daily limits on multiple capable models (like o4-mini) to a super restrictive cap on GPT-5 and a tiny weekly limit on the ‘thinking’ model? Math ain’t mathin.”
The workaround currently circulating – explicitly instructing the model to “think longer” within prompts – feels like a band-aid solution for a fundamental issue. It places the burden on users to compensate for limitations imposed by the platform itself.
The Future of AI Access: Balancing Demand and Value
OpenAI faces a delicate balancing act. The company needs to manage the immense global demand for its advanced AI models while maintaining subscriber satisfaction. The current limitations on GPT-5 access, particularly for Plus users, risk alienating the very customers who have supported OpenAI’s development. This situation highlights a growing tension in the AI landscape: how to monetize powerful AI tools without restricting access to those who value them most.
One potential solution lies in dynamic pricing and tiered access levels that more accurately reflect usage patterns and computational costs. Another could involve exploring alternative models for “Thinking” capabilities, perhaps offering dedicated, paid add-ons for users requiring extensive reasoning power. The current model, however, feels like a downgrade for a significant segment of the user base.
The long-term implications extend beyond ChatGPT. As AI models become increasingly sophisticated, the question of equitable access will become even more critical. Will advanced AI remain a privilege for those who can afford it, or will it become a widely accessible tool for innovation and problem-solving? The answer will shape the future of technology and its impact on society. The World Economic Forum has published extensive research on the ethical considerations of AI access, offering valuable insights into this complex issue.
What are your thoughts on OpenAI’s new GPT-5 access model? Do you think the limitations are justified, or are they a step in the wrong direction? Share your perspective in the comments below!
1 comment
So these limits for Plus users require a person to literally say everything in one message or sit and tabulate how many back and forth messages you have with the AI at a session.
It’s 80 messages in 3 hours for Plus users. And by the way there is no built in way to see how many messages (in essence credits) you have spent at any one time. You only know you’ve gone over when you get this bs message “You’ve gone over your limit. You have access x hours later.”
What the company is failing to account for is that the AI is designed to have a back and forth with a user to communicate and function. Who will keep track of their precious 80 messages to the AI?
What if the AI makes a mistake (happens frequently) through the no fault of the user? And now you have to spend basically one of your 80 credits to fix a mistake not due to a user error but an AI error?
How is this reasonable? For the amount of money they are charging every month, Plus subscribers are paying a huge amount for a very limited amount of access. This is literally highway robbery on the backs of its core customers.
I hope the company comes up with a better response.
I for one have a very serious illness. I am using the AI not for kicks but to help plan reasonable next steps to fight my illness. This is literally my life here. And this new tier system is unconscionable.
Without a reasonable amount of access, people all over the world will be priced out and it will be the haves vs have nots. Classism at its worst. The company should be ashamed.
I am truly shocked. I hope someone has the good sense to correct this.