GPT5 launched on August 7, 2025, across all OpenAI ChatGPT tiers, bringing major improvements in reasoning, accuracy, and accessibility. The model now handles up to 256,000 tokens in a single conversation, offers reduced hallucinations, and delivers better factual reliability. It integrates with Gmail and Google Calendar, supports voice interactions, and allows customizable personalities. Sam Altman calls it “a PhD-level expert” across coding, science, finance, health, and creativity.
The biggest functional change is automatic routing to the right reasoning mode, removing the need for users to pick between fast or deep-thinking versions. OpenAI has also made the base model available to free users, while advanced features are unlocked in Plus and Pro tiers.
Controversy came fast, with many upset that GPT-4o was removed in the update. Users on Reddit, Inc. and X accused OpenAI of stripping GPT-4o’s personality and began cancelling subscriptions until it returns. Sam Altman acknowledged the backlash and said he would consider restoring GPT-4o for Plus subscribers.
Critics also flagged GPT-5’s launch demo, where the model made simple errors, raising questions about whether it is ready for full-scale use. Some analysts praise it as a strong step forward, while others see it as evolutionary, not revolutionary.
Read more: Onic Offers Free GPT-4o Premium to All Users