After several delays, OpenAI says its next foundational model, GPT-5, could launch as early as next month, The Verge and Axios report. The timing is tentative, but there are indications that the release is imminent: Last week, OpenAI CEO Sam Altman teased on social media that it was coming “soon,” and testers and “red teams” are gauging its capabilities and security. GPT-5 is expected to merge traditional and reasoning models like o3, with mini and nano versions also available through its API.
OpenAI is preparing to launch its most powerful artificial intelligence model yet—GPT-5—as early as August, according to The Verge, citing sources familiar with the company’s internal timeline.
The release, if it happens as expected, will mark a critical leap in generative AI development and set the tone for the next stage of competition among global AI powerhouses.
The model has been under intense internal testing for months. Back in May, Microsoft engineers were reported to be provisioning significant server capacity in anticipation of a possible launch, prompting speculation about a release timeline that has since shifted slightly. Sources say that OpenAI has now entered the final phases of testing and refinement, with the infrastructure upgrades and technical evaluations largely in place.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026): big discounts for early bird.
Tekedia AI in Business Masterclass opens registrations.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab: From Technical Design to Deployment (next edition begins Jan 24 2026).
OpenAI CEO Sam Altman has been steadily teasing GPT-5’s capabilities. In a recent podcast appearance with comedian Theo Von, Altman shared how the model solved a question he struggled with, prompting him to say he felt “useless relative to the AI.” He also confirmed publicly via a post on X that the model is coming soon, adding to growing anticipation across the developer and enterprise ecosystem.
One key element of the launch is the company’s planned release of mini and nano variants of GPT-5, designed to serve a range of needs from lightweight consumer applications to enterprise-grade deployments. These models will be offered via OpenAI’s API and platform, positioning the company to expand its reach across varied sectors, including software development, creative industries, customer service, finance, and education.
Unlike prior rollouts—such as GPT-4o, which introduced real-time multimodal capabilities—GPT-5 is expected to integrate a broader range of functionalities into a single, cohesive system. Altman has described the model as “a system that integrates a lot of our technology,” with internal discussions suggesting GPT-5 will combine memory, reasoning, multimodal input comprehension, and potentially real-time learning into a unified AI agent.
This bundling strategy reflects OpenAI’s intent to streamline user interaction with its tools by reducing fragmentation across different versions. The goal, according to developers close to the company, is to simplify deployment and ensure consistency in performance and behavior across devices and platforms.
The upcoming release is also seen as OpenAI’s strategic response to intensifying pressure from rivals like Google’s Gemini, Anthropic’s Claude, Meta’s Llama series, and Mistral’s open-weight models. As these companies race to dominate both foundational AI and consumer-facing interfaces, GPT-5 could determine whether OpenAI maintains its current leadership position or begins to cede ground to competitors offering faster, cheaper, or more open alternatives.
Beyond raw performance, OpenAI’s model strategy also carries significant commercial weight. Microsoft, a major investor in OpenAI, is expected to embed GPT-5 across its Azure cloud platform, Copilot products, and enterprise offerings. The integration could push further adoption of AI-powered workflows within Microsoft’s global client base, from small businesses to Fortune 500 firms.
As the launch window draws closer, developers and researchers are paying close attention to how GPT-5 will handle long-form reasoning, factual consistency, and safety guardrails—key issues that have dogged even the most advanced models. OpenAI has faced criticism over hallucinations and content control, and expectations are high that GPT-5 will demonstrate meaningful progress on these fronts.
For OpenAI, this is more than a product upgrade—it’s a litmus test for whether generative AI can transition from experimental hype to mission-critical infrastructure. GPT-4 was criticized for hallucinating, after users had expected that the model would not come with the shortfall. Even Altman said he was surprised at the model’s level of hallucination. GPT-5 is thus expected to be a better version, with many of the needed corrections.