Users Are Disappointed with ChatGPT 5 and Call It Horrible

Users Are Not Happy with ChatGPT 5 and Call It Horrible

Less than 24 hours after OpenAI rolled out its highly anticipated GPT-5, initial feedback from users suggests that the launch might not have hit the mark. Far from universal praise, a wave of disappointment has swept across user communities.

The Launch and Immediate Fallout

During a one-hour livestream event, OpenAI CEO Sam Altman and his team showcased GPT-5, touting its advancements in reasoning, multi-step processing, and enhanced personalization. The new model was presented as a “PhD-level” upgrade, promising a significant leap forward in AI capabilities from the previous GPT-4o series.

However, the enthusiasm from OpenAI quickly met a wall of criticism online. Within hours of the reveal, social media platforms, particularly Reddit, became a hotbed of dissatisfaction. A Reddit thread titled “GPT-5 is horrible” rapidly accumulated nearly 3,000 upvotes and over 1,200 comments, with countless users voicing their frustration and even demanding the return of older, more reliable models like GPT-4.

Poor Performance, Slower Speeds, and Hidden Costs

The core complaints are consistent: many users report a noticeable decline in response quality, slower output, and new limitations, especially for paid subscribers. GPT-5’s new “Thinking” mode, designed for complex reasoning, is capped at a meager 200 messages per week for ChatGPT Plus users. To make matters worse, Plus subscribers now have fewer model choices, as OpenAI has consolidated its offerings under GPT-5, claiming the model can supposedly self-select reasoning depth as needed.

Some users have accurately described the situation as “shrinkflation,” arguing that while GPT-5 might score higher on internal benchmarks, its practical usability and perceived value feel diminished. A common sentiment is frustration that previous, more dependable models were removed entirely instead of being offered alongside the new upgrade, forcing users onto a version they don’t prefer.

Were Expectations Too High?

The intense criticism follows a significant marketing build-up. Sam Altman himself teased GPT-5’s debut with dramatic, Star Wars-inspired social media posts, hinting at a truly game-changing innovation. While GPT-5 might show measurable performance improvements in controlled testing environments, many users feel the update is merely incremental rather than the revolutionary leap they were led to expect. The consensus among the discontented is that GPT-5 pales in comparison to the groundbreaking impact of ChatGPT’s initial release back in 2022. It appears OpenAI now faces the challenge of appeasing a user base that feels let down by its latest flagship AI.