OpenAI's most powerful AI model is here and free for everyone

In July, OpenAI CEO Sam Altman confirmed that the corporate’s extremely anticipated new AI model, GPT-5, was nearing launch. A report later claimed that the model was on monitor for an August launch. This week, the corporate posted on X (previously Twitter) to inform customers {that a} livestream would occur on Thursday, August 7. It seems that the livestream ended up being the announcement for the rollout of GPT-5.

In the present day, OpenAI is beginning to roll out GPT-5 to all of its customers. In a slightly prolonged blog post, the corporate reveals that the model is obtainable right this moment for Plus, Professional, Crew, and free customers. Nevertheless, Enterprise and Edu subscribers must wait one week earlier than gaining entry. There are three new fashions in whole: GPT-5, GPT-5 mini, and GPT-5 Professional.

OpenAI notes that Plus subscribers will be capable to use GPT-5 considerably greater than free customers. When free customers attain their restrict, they’ll be transitioned to a smaller, sooner model referred to as GPT-5 mini. In the meantime, Professional members will be capable to use the model as a lot as they need and may have unique entry to GPT-5 Professional, a model with prolonged reasoning capabilities.

In keeping with the corporate, GPT-5 is a reasoning model, which might be the primary of its fashions to supply this function by default. When you’re unfamiliar with the time period “reasoning model,” it signifies that the system can break down complicated issues into smaller components and work its solution to a solution step-by-step. Beforehand, you would need to get ChatGPT to make use of one OpenAI’s reasoning fashions by choosing the “Suppose Longer” choice within the immediate bar.

Along with being a reasoning model, the AI is mentioned to boast enhancements throughout the board in comparison with earlier GPT fashions. Particularly, this AI model affords higher accuracy, experiences fewer hallucinations, solutions questions sooner, is safer, and is extra trustworthy. The corporate factors to creating the largest enhancements with coding, well being, and writing.

For instance, the corporate claims that GPT-5 responses are 45% much less more likely to include a factual error than GPT‑4o, and that quantity climbs as much as 80% in comparison with OpenAI o3. By way of honesty, it gained’t lie about finishing a process, be overly assured about an unsure reply, and sycophancy has been minimized. It additionally features a new function referred to as Protected Completions, the place GPT-5 will attempt to give the most useful reply inside security confines. Beforehand, OpenAI used a special method the place the model would both adjust to or refuse a immediate.

On high of the GPT-5 information, OpenAI introduced some new customizations are rolling out for ChatGPT. The chatbot may have 4 new personalities that will probably be obtainable initially for textual content chat, with Voice coming later. These 4 personalities embody Cynic, Robotic, Listener, and Nerd. They’re obtainable as a part of a analysis preview and might be modified or disabled at any time.