After literally years of hype, OpenAI finally has The company has officially launched its new lineup of large-language models (LLMs), which includes different-sized variants GPT-5. This is the long-awaited successor to GPT-4, which was released in March 2023, almost 2.5 years ago.
To meet the needs of different users, the company is releasing a number of versions – GPT-5 GPT-5 Nano and GPT-5Pro – to suit their varying needs in terms speed, cost and computational depth.
GPT-5 is soon to be exclusively powering ChatGPTandreplace all other models going forward. ChatGPT Pro ($200) subscribers can still choose older models for the next sixty days.
According to rumors and reports from OpenAI, the previous system that had users select the underlying model that powered ChatGPT has been replaced with a routerwhich decides whether to use a special GPT-5 thinking mode with “deeper reason” that takes longer to answer harder queries or the regular GPT-5 models or mini models for simple queries.
AI Scaling Has Its Limits.
Power limits, rising token costs, inference delays, and other factors are reshaping enterprise AI. Join our exclusive event to learn how top teams are: https://bit.ly/4mwGngO
In the API, the three reasoning-focused models — GPT-5, GPT-5 mini, and GPT-5 nano — are available as gpt-5, gpt-5-mini, and gpt-5-nano, respectively. GPT-5 Pro, which is used to power ChatGPT only for Pro-tier subscribers, is not accessible via API.
GPT-5 was released just days after OpenAI launched a new set of open source LLMs called GPT-oss, which can be customized and used by individuals and developers offline on consumer devices such as PCs/Mac desktops or laptops.
What’s most important is not what GPT-5 does, but what it doesn’t do: AGI (artificial general intelligence), OpenAI’s LLMs. The stated goal is to create an autonomous AI system capable of outperforming humans in the most valuable economic work.
OpenAI’s declaration of AGI will have a material impact on business, whether or not you personally believe that such a system would be desirable or possible. Wired previously reported that there was a clause in OpenAI’s contract with Microsoft which allows OpenAI to charge Microsoft for access to their newest models or block it from accessing OpenAI’s models if OpenAI’s board determines that the company has achieved AGI, or generates over $100 billion in profits.
However, this is not the situation today. As CEO Sam Altman stated, flanked with other OpenAI staffers, on an embargoed, video call with journalists last night: ” The way that most people define AGI, there’s still something quite important that we’re missing – many things that are important, in fact — But one big one is that a model continuously learns as it deploys, and GPT-5 doesn Will it change the Microsoft negotiations?
An OpenAI spokesperson replied via email:
GPT-5 represents a significant step towards AGI, as it shows improvements in reasoning and generalization. AGI is a term that is not well defined and can mean different things to different people.Although GPT-5 meets early criteria for AGI it has not yet reached the threshold of fully-human-level AGI.There remain key limitations in areas such as persistent memory, autonomy and adaptability between tasks. We are focused on improving these capabilities in a safe manner, not on speculating about specific timelines.Yet OpenAI benchmark results show GPT-5 is close to performing as well, or even exceeding, the average expert performanceacross law, logistics and sales.
OpenAI writes that GPT-5 performs as well or better than experts on average in about half of the cases. It also outperforms OpenAI o3 & ChatGPT Agent.
Why use GPT-5 What does GPT-5 have to offer in the face of so many alternative models from OpenAI, and a growing number of competitors, including Chinese startups that offer powerful open-source models? Altman described the capability leap as more than incremental. Altman compared the experience with GPT-5 to upgrading a pixelated screen to a retina display — something users don’t want back from. Altman said that GPT-3 was like talking to high school students. “GPT-4 felt like talking to a college student.”GPT-5 feels like talking to an expert at PhD level in your pocket.
One of the most impressive abilities demonstrated for reporters during the embargoed calls was the ability to generate the code for a working web application using a single prompt.In this case, the app was a French Language Learning App with a built-in game that displayed English-to French phrases every time a user guided a mouse to collect cheese. The prompt was also only one paragraph.
Altman said : “This concept of software on-demand will be a defining feature of the new GPT-5 era.”
GPT-5’s advantage in creating games, apps, or other software from prompts is speed. It produced this demo app within minutes. It also seems to offer completeness with few bugs and a playable experience from a single request, as the developers say.
GPT-5 is available OpenAI made the model[19659032[19659032
Performance improvements show up across key academic and real-world benchmarks. In coding, GPT-5 sets new state-of-the-art results on SWE-Bench Verified (74.9%) and Aider Polyglot (88%).
Perhaps most incredibly, on Humanity’s Last Exam — a newish benchmark of 2,500 extremely difficult tasks for programs — GPT-5 pro achieves a record-high 42%, blowing away the competition and all prior OpenAI models except the new ChatGPT agent unveiled last month that controls its own computer and cursor like a human.
On writing tasks, GPT-5 adapts more smoothly to tone, context, and user intent.It is better at maintaining coherence, structuring information clearly, and completing complex writing assignments.
The improvements are not just technical — OpenAI’s team emphasized how GPT-5 feels more natural and humanlike in conversation.
Health-related use cases have also been enhanced. While OpenAI continues to caution that ChatGPT is not a replacement for medical professionals, GPT-5 is more proactive about flagging concerns, helping users interpret medical results, and guiding them through preparing for appointments or evaluating options. The system also adjusts answers based on user location, background knowledge, and context — leading to safer and more personalized assistance.
One of the most significant updates is in safe completions a new system that helps GPT-5 avoid abrupt refusals or unsafe outputs.
Instead of declining queries outright, GPT-5 aims to provide the most helpful response within its safety boundaries and explains when it cannot assist — a change that dramatically reduces unnecessary denials while maintaining trustworthiness.
GPT-5 is also a major upgrade for developers working on agentic systems and tool-assisted workflows. OpenAI has introduced a suite of developer-friendly controls in the GPT-5 API, including:
Free-form function calling – Tools can now accept raw strings such as SQL queries or shell commands, without requiring JSON structure.
Reasoning effort control – Developers can toggle between rapid responses and deeper analytical processing depending on the task.
Verbosity control– A new parameter allows users to select whether responses are brief, standard, or detailed.
Structured outputs with grammar constraints – Developers can now guide outputs using custom grammars or regular expressions.
Tool call preambles – GPT-5 can now explain its reasoning before using tools or making external requests.
The Advanced Voice mode allows users to adjust the tone and delivery of their voice. Voice will be available to all users and in custom GPTs.
OpenAI will retire its older “Standard Voice Mode”and transition to a unified experience in 30 days.
GPT-5’s safer design, robust reasoning, expanded developer tools, and wide user access reflect a maturing AI eco-system that is moving closer to real-world application on a global level. OpenAI’s strategy this time is less flashy and more integrated. GPT-5 doesn’t have to be a separate product that users must seek out. It’s already there, powering millions of tools, making them more capable and smarter, and unlocking new use cases for developers.
VB Daily provides daily insights on business use-cases
Want to impress your boss? VB Daily can help. We provide you with the inside scoop about what companies are doing to maximize ROI, from regulatory changes to practical deployments.
Read our privacy policy
Thank you for subscribing. Click here to view more VB Newsletters.
