OpenAI’s latest model, GPT-5, has had a tumultuous launch. Despite much anticipation, it has faced significant criticism. Users reported that GPT-5 struggled with simple problems, including math and algebra questions that prior models and competitors could solve effortlessly. It mistakenly judged errors in OpenAI’s own presentation data and underperformed on more complex tasks compared to other AI models like Anthropic’s Claude Opus 4.1.
OpenAI’s decision to phase out older models, while continuing to use them internally, has frustrated many ChatGPT users. Security reports also highlighted gaps in GPT-5’s safety measures. Initial feedback on the model has been tepid, with many users expressing disappointment.
Furthermore, new offerings from competitors such as Alibaba’s Qwen 3, which allows for more extensive interaction, present a growing challenge for OpenAI. Some experts believe opinion could shift as users adapt to GPT-5’s capabilities, but the initial release has not met expectations. As OpenAI remains under financial pressure, the success of GPT-5 is crucial to their future prospects.