ChatGPT-5: New Model, Features, and Availability

OpenAI announces that GPT-5 will have millions of users in the very first month thanks to its existing user base

With over 180 million active monthly users registered on ChatGPT at the end of 2024, OpenAI expects that the release of GPT-5, as soon as it becomes available, will lead to a new peak in sign-ups and usage, probably surpassing 200 million global users already in the first month; the rapid spread of previous versions suggests that the adoption of the new model will be even faster, also because access will be granted both to free plan users and those who choose Plus and Pro subscriptions.

GPT-5 promises a much wider memory and superior text comprehension compared to GPT-4

While GPT-4 manages on average a context window of 128,000 tokens, equivalent to over 300 pages of continuous text, rumors about GPT-5 suggest an even longer memory, with the ability to process larger files and keep track of very detailed conversations; according to OpenAI data, the jump from GPT-3 to GPT-4 already resulted in a 50% increase in accuracy on complex questions, and internal tests on early GPT-5 versions show a further increase in accuracy and consistency, which will be measurable as soon as the model is made public.

The launch of GPT-5 should include a slower but more precise reasoning mode called “Strawberry”

According to previews shared by OpenAI and picked up by tech magazines such as TechRadar, the “Strawberry” function will allow GPT-5 to take up to 20 extra seconds to analyze difficult requests, providing more thoughtful answers and reducing errors; it’s estimated that this option, tested on a sample of 50,000 complex queries during the beta phase, has already reduced incoherent or approximate answers by 30% compared to GPT-4; for advanced users and those working on research projects, this innovation could be crucial in handling sophisticated tasks.

The new model will also be accessible from the free plan, with some technical limitations compared to Plus and Pro

According to official statements from OpenAI, even users who access ChatGPT without a subscription will be able to use GPT-5, but with lower response speed and usage priority compared to subscribers; currently, about 10% of the total user base chooses a paid subscription, but OpenAI expects that with the arrival of GPT-5 this percentage could rise to 15-20%, driven by new advanced features and an overall productivity increase estimated between 18% and 22% in the first months of adoption.

The applications of GPT-5 will extend to more and more industries thanks to multimodal support

While GPT-4 is already used in industries such as education, healthcare, finance, and customer service, version 5, with its ability to handle text, images, voice, and potentially even video, is set to multiply employment opportunities; according to a survey conducted by McKinsey at the start of 2025 on over 1,000 companies already adopting AI, 72% plan to integrate OpenAI models in at least two new productive or service processes by the end of the year, with an estimated impact on operating cost savings between 12% and 19%.

Media coverage and user expectations continue to grow week by week

Since the beginning of 2025, mentions of GPT-5 on major social networks and online media have increased by 350% compared to the same period last year, reaching, in June alone, over 4 million posts and articles worldwide; the volume of Google searches linked to the query “ChatGPT-5” has grown by 240% compared to January, signaling that the public, professionals, and companies are increasingly attentive to updates and official announcements; OpenAI is leveraging this wave of interest to collect feedback and prepare a global launch campaign that can involve all strategic markets.

The adoption of scalable architectures and parameter optimization will be the key to improving GPT-5’s computational efficiency

To respond to the growing demand for computing capacity, OpenAI has had to invest in new optimization strategies for neural architectures, choosing not only to increase the number of model parameters compared to previous versions, but also to implement pruning and quantization techniques that allow performance and energy sustainability to be balanced; in this context, the decision to adopt modular and scalable architectures, combined with new distributed training systems, aims to drastically reduce training times and infrastructure costs, thus offering companies and researchers a tool that promises levels of customization and adaptability never before reached; the real challenge, however, will be to make this efficiency accessible even to those who do not have the resources of tech giants, ensuring democratization of artificial intelligence without sacrificing answer quality or depth of reasoning.

Data governance and training transparency will define the level of trust the market will place in GPT-5

With the introduction of GPT-5, the pressure from regulators, governments, and public opinion for OpenAI to adopt more transparent policies regarding training data provenance, bias management, and documentation of algorithmic choices has increased exponentially; the new generation of models will therefore need to integrate mechanisms for internal auditing, decision logs, and accessible reporting, so that companies and public administrations can verify the correctness and reliability of the answers obtained at any time; from this perspective, OpenAI’s ability to share details on datasets, sources, and anonymization processes will become a crucial competitive factor, not only to differentiate itself from competitors, but above all to gain the trust of highly regulated markets such as healthcare, finance, and the public sector.

The ethical implications and manipulation risks will be more sophisticated and require new forms of active monitoring

The exponential increase in power and flexibility of GPT-5 inevitably also entails growing risks linked to malicious use, misinformation, and psychological manipulation of users, who will now be able to interact with models that are ever more persuasive and harder to distinguish from a human interlocutor; OpenAI will therefore have to implement advanced systems for watermarking responses, real-time semantic filters, and algorithms for detecting behavioral anomalies, capable of identifying abuse patterns even at scale; only through proactive commitment and collaborations with independent research bodies will it be possible to prevent new forms of attack or exploitation, thus ensuring that the evolution of artificial intelligence remains aligned with democratic values, protection of individuals, and respect for global regulations.

The arrival of GPT-5 may change geopolitical balances and stimulate a new global race for advanced artificial intelligence

With the increasing centrality of AI in the decision-making processes of governments, strategic companies, and critical infrastructures, the launch of GPT-5 could represent a turning point for the balance of global powers, favoring those countries and companies able to integrate the new features into their workflows first; competition between the United States, China, the European Union, and other emerging economies will also hinge on the ability to access advanced language models, leveraging proprietary algorithms for defense, cybersecurity, and strategic information, while the lack of regulated and balanced access risks accentuating digital divides and instability in international markets; technological diplomacy and multilateral cooperation will thus become essential elements to manage the risks of a digital arms race that—without a clear regulatory framework—could slip out of the hands of the actors involved.

The skills required of future professionals will be redefined by the integration of GPT-5 into business processes, creating new opportunities and challenges

The adoption of GPT-5, which will be able to automate complex tasks such as the drafting of strategic reports, predictive analysis, or the integrated management of multimodal data, will entail a radical transformation of the skills required of knowledge workers, who will have to develop new digital literacy not only aimed at the effective use of AI, but also at its critical supervision and ethical interaction with automated systems; thus, hybrid professional roles will emerge, able to interact with advanced algorithms, assess result accuracy, and promptly identify risks or biases, while organizations will have to invest in ongoing training and policy updates to avoid both the risk of obsolete skills and that of delegating too much to machines at the expense of creativity and human judgment.

The future evolution of AI models will depend on collaboration among the open source community, large private players, and new institutional actors

While OpenAI and major companies in the sector will continue to push innovation with proprietary investments and research, the success and impact of GPT-5 will be strongly influenced by the ability to foster an open ecosystem, where the open source community, universities, and new standardization bodies can contribute to the definition of best practices, interoperability frameworks, and shared solutions for data security and governance; only through active collaboration between public and private sectors, fueled by real transparency in technological choices and widespread dissemination of knowledge, will it be possible to guide the development of future generations of artificial intelligence toward inclusivity, sustainability, and genuine social utility.

_______

FOLLOW US FOR MORE.

DATA STUDIOS

datastudios.org