#ICYMI: OpenAI launches GPT-5, claims to be its "smartest and fastest ever"

The new AI features a model that answers most questions, a deeper reasoning model for harder problems, among other things.

Screenshot from the ChatGPT website

Smarter and more widely useful large language model

OpenAI has launched GPT-5, its latest large language model (LLM) that features more focus on enhanced reasoning capabilities and a more unified, agent-like system.

According to the developer, GPT‑5 is a unified system that has the capacity to answer most questions, a deeper reasoning model (GPT‑5 thinking) for harder problems, and a real‑time router that quickly decides which to use based on conversation type, complexity, tool needs, and the explicit intent.

The router is continuously trained on real signals, including when users switch models, preference rates for responses, and measured correctness, improving over time, OpenAI said. Once usage limits are reached, a mini version of each model handles remaining queries.

The company aims to integrate these capabilities into a single model in the future.

In GPT-5, OpenAI said the new LLM answers questions more quickly and is more useful for real-world queries. Improvements were made to reduce hallucinations, improve instruction following, and minimize sycophancy, while leveling up performance in three of ChatGPT’s most common uses: writing, coding, and health.

When it comes to coding, GPT‑5 shows particular improvements in complex front‑end generation and debugging larger repositories.

It can often create beautiful and responsive websites, apps, and games with an eye for aesthetic sensibility in just one prompt, intuitively and tastefully turning ideas into reality. Early testers also noted its design choices, with a much better understanding of things like spacing, typography, and white space, OpenAI said.

When it comes to writing, GPT‑5 is seen to be more reliable when it comes to those that involve structural ambiguity. The improvements mean that the latest LLM is more helpful in drafting and editing reports, emails, memos, and more.

If users have health-related questions, GPT-5 scores significantly higher than any previous model on HealthBench⁠.

Compared to previous models, it acts more like an active thought partner, proactively flagging potential concerns and asking questions to give more helpful answers, OpenAI said.

GPT‑5 also now provides more precise and reliable responses, adapting to the user’s context, knowledge level, and geography, enabling it to provide safer and more helpful responses in a wide range of scenarios.

OpenAI emphasized that ChatGPT does not replace a medical professional. This only serves as a partner to help users understand results, ask the right questions in the time they have with providers, and weigh options as they make decisions.

More information about the new GPT‑5 can be accessed here.

Source: OpenAI