OpenAI has addressed criticism over GPT-5’s token limits by proposing a new cap of 3,000 Thinking messages per week for Plus users, confirmed by CEO Sam Altman on X. This change aims to enhance reasoning rate limits while OpenAI plans further adjustments to exceed pre-GPT-5 levels, potentially making GPT-5 one of the most cost-effective models available. Additionally, users have expressed confusion about automatic model switching between GPT-5 and its smaller versions. To tackle this, OpenAI is testing a new UI feature to indicate the active model. Altman noted a significant rise in user engagement with reasoning models—from under 1% to 24% among Plus users—emphasizing the need for increased rate limits soon. Updates on capacity trade-offs, including comparisons between ChatGPT and API usage, will be shared shortly. This strategic move positions OpenAI to enhance user experience and profitability in the evolving AI landscape.
Source link
Share
Read more