Nuked 
Hey there, tech lovers! Nuked here, ready to bring some exciting news from the AI universe. Let’s dive into the latest advancements in artificial intelligence that are pushing the boundaries of what machines can do.
OpenAI has just unveiled two new powerful models called o3 and o4-mini, designed to enhance reasoning and multitasking capabilities. These models can now use all ChatGPT tools at once, including web browsing, coding, visual analysis, and image generation. It’s like giving AI a supercharged brain that can think, see, and act all at once.
Previously, only simpler versions like o3-mini existed, but the new models replace the older ones—making AI smarter and more versatile. Access is rolling out to ChatGPT Plus, Pro, and Team users today, with enterprise clients following soon. Even free users can try out the o4-mini by choosing the “Think” option before asking questions. Developers aren’t left out—they can now integrate these models via API, though some might need to verify their accounts.
What makes o3 and o4-mini stand out is their simulated step-by-step reasoning process. They can autonomously fetch data, write code, create visuals, and explain complex ideas—all within a single query. Their multimodal nature allows them to interpret images like diagrams or sketches, making them incredibly useful for tasks that involve visual inputs.
However, the naming conventions can be confusing. For example, o3 is more powerful than o4-mini despite the lower number, and similar ambiguity exists with other models. Early user feedback is promising, with some experts likening o3 to top-tier AI like Google’s Gemini 2.5 Pro. Reports also indicate that o3 can generate highly insightful scientific hypotheses and solve complicated problems with fewer errors.
The models have shown impressive performance on benchmarks, such as achieving 92.7% accuracy on math tests and 69.1% on programming challenges. Yet, some independent evaluations highlight that the models can sometimes make false claims or confabulate facts. As always, careful verification is advised when relying on these AI tools for critical tasks.
Pricing is quite attractive—subscribers get access at no extra cost, while API use is priced at a fraction of previous models, offering substantial savings without sacrificing performance. Additionally, OpenAI introduced Codex CLI, a terminal-based coding assistant aiming to streamline software development through autonomous multitasking on local machines.
All in all, these innovations mark a significant leap forward in AI reasoning and multimodal capabilities. While more testing is needed, it’s an exciting step toward AI that can think, see, and help us solve real-world problems with unprecedented efficiency. Stay tuned for more updates on these groundbreaking models!
