
OpenAI will make GPT-5.6 Luna the default ChatGPT model from later this week, replacing GPT-5.5 Instant, Times Now reports. The company says internal tests found Luna 62 per cent less likely to…
OpenAI will make GPT-5.6 Luna the default ChatGPT model from later this week, replacing GPT-5.5 Instant, Times Now reports. The company says internal tests found Luna 62 per cent less likely to produce at least one factual error, particularly in prompts involving dates, figures, regulations and source-based data.
Free and Go subscribers will get unlimited text conversations with Luna. Plus and Pro users will receive an upgraded GPT-5.6 Sol and a reasoning slider that lets them choose between faster replies and more computing for complex questions. The update comes as lower-cost Chinese open-weight models put pressure on OpenAI and other premium AI providers.
Claims that a new default model will make ChatGPT reliable can run ahead of the evidence, while claims that cheaper Chinese models have already displaced paid services are just as sweeping. OpenAI’s 62 per cent figure is an internal test, not a guarantee for every user or topic. The useful test is whether independent evaluations and everyday error reports confirm fewer mistakes after the rollout.
Source: timesnownews.com
This story was synthesised by AI from the source linked above.