OpenAI Brings Back GPT-4o, Boosts GPT-5 Limits for Plus Users

After the rollout of GPT-5, which is now accessible to all users, the company is implementing notable changes. These include reintroducing the well-liked GPT-4o model and increasing GPT-5's usage limits. ChatGPT Plus subscribers can now select between GPT-5, GPT-5 Thinking, and GPT-4o models.

Sam Altman announced that Plus users can continue using the GPT-4o model. He stated, "We will watch usage as we think about how long to offer legacy models for." This decision comes after feedback from users who found GPT-5 less engaging compared to previous versions.

Increased Usage Limits for Subscribers

Altman also confirmed a significant increase in rate limits for reasoning for ChatGPT Plus users. "Today we are significantly increasing rate limits for reasoning for ChatGPT Plus users, and all model-class limits will shortly be higher than they were before GPT-5," he mentioned. Additionally, a UI change will soon indicate which model is in use.

For those unaware, free users have access only to GPT-5 with capped usage. Once this limit is reached, the chatbot automatically switches to a lighter version called GPT-5 mini. This ensures that even free users can continue using the service without interruption.

Options for Pro and Team Customers

Pro and Team customers enjoy full access to both the standard GPT-5 and the research-grade GPT-5 Pro model. Altman shared that they are considering offering a limited number of monthly queries on the GPT-5 Pro model to Plus subscribers as a trial. "We are considering giving a (very) small number of GPT-5 pro queries each month to plus subscribers so they can try it out! I like it too," he said.

Despite some criticism from nearly 5,000 Reddit users who described GPT-5 as a "downgrade," these changes aim to address user concerns by providing more options and flexibility. The company hopes these adjustments will improve user satisfaction by catering to diverse preferences.

The company is also exploring ways to accommodate requests from those willing to pay $1,000 monthly for double input tokens. Altman remarked on this possibility: "But yeah if you wanna pay us $1k a month for 2x the input tokens feels like we should find a way to make that happen..."

The introduction of these changes reflects an effort to balance innovation with user feedback. By offering multiple models and adjusting usage limits, the company aims to enhance user experience across different subscription tiers.