OpenAI launches its most advanced AI model Gpt-5
OpenAI has launched its most advanced AI model till date, GPT-5. This is an advanced version of ChatGPT’s previous model GPT-4o. The company claims that coding, medical research and writing have been made more advanced in this tool. It will also be able to answer complex mathematical questions. Also, it can do many things including visual perception.
What is special in GPT-5?
OpenAI told about this advanced AI tool in its blog post that it is currently being rolled out for all users. This means that it will be available to all users, free and paid. Free users along with GPT Plus, Pro and Team subscription will be able to use it from now on. At the same time, enterprise and education users will get access to it from next week.
Paid users will also get access to Codex CLI in this tool, which will be able to perform many advanced operations. Apart from this, paid users will also get access to GPT-5 Thinking in it. The company has confirmed that this new model will replace GPT-4o, GPT o3, o4 Mini, GPT 4.1 and GPT 4.5. However, users will be able to use the old GPT models by selecting them from the model picker tool. It also has the ability to generate quick responses from reasoning.
GPT-5 will have four options including Cynic, Robot, Listener and Nerd. These can match the communication style of the users, due to which the conversation will be even better. OpenAI claims that this AI tool has been tested for more than 5 thousand hours. Strong safety tools have been added to it based on the testing feedback of internal employees.
How much is the plan?
Paid users will get access to GPT-5 in the Pro plan, for which users will have to spend $200 i.e. about Rs 17,500 per month. After this, they will be able to get unlimited access to this advanced tool. The company told about this advanced AI tool that it has a unified LLM, which will work to improve coding, health, math and visual perception to writing.
Talking about the benchmarking of GPT-5, this tool has got a score of 74.9 percent on SWE, which is much higher than GPT-4.1 which got a score of only 54.6 percent here.
