GPT-5 backlash: ChatGPT Plus users say the upgrade feels like a downgrade

Since the launch of OpenAI's GPT-5 model last week, a wave of discontent has emerged among users regarding its performance compared to earlier versions. Many users have criticized GPT-5 for its tendency to provide concise and emotionally detached responses, raising concerns that the update may be more about cost-cutting than improving user experience. Compounding these frustrations was OpenAI's decision to remove access to previous models like GPT-4o, o3, and o4-mini as GPT-5 became available. Although subscribers to ChatGPT Pro and Team could still utilize these older models through legacy settings, Free and Plus tier users found themselves cut off from valuable resources. While the message limits for GPT-5 on ChatGPT Plus were largely unchanged from those of GPT-4o, many users felt that the new model represented a net loss when compared to previous capabilities. Under the new system, Plus users are allowed 80 messages every three hours for the standard model, and 200 messages per week for the GPT-5 Thinking model. In contrast, before the update, they had access to multiple models with significantly higher limits, including 700 messages per week for o4-mini high and 2,100 for o4-mini. Additionally, the context window for Plus subscribers remained at 32,000 tokens (approximately 24,000 words), while higher-tier users enjoyed a much larger context window of 128,000 tokens (around 96,000 words). This limitation hinders the AI's ability to manage and recall longer conversations effectively. The launch of GPT-5 sparked intense debate on social media, particularly within ChatGPT and OpenAI communities, where numerous users threatened to cancel their Plus subscriptions. In response to user backlash, OpenAI took immediate steps to address these concerns. The company increased the rate limit for GPT-5 on the Plus plan to 160 messages every three hours and restored access to GPT-4o for Plus users through legacy model settings. Furthermore, CEO Sam Altman announced plans to enhance the limits for the GPT-5 Thinking model, aiming to provide users with up to 1,000 messages per week. These adjustments reflect OpenAI's commitment to improving user satisfaction amid growing criticism.