OpenAI Unveils ChatGPT 5: ‘Like Talking to an Expert Who Knows Everything,’ Says Sam Altman

OpenAI Unveils ChatGPT 5: ‘Like Talking to an Expert Who Knows Everything,’ Says Sam Altman
X
OpenAI launches ChatGPT 5, combining advanced reasoning, reduced errors, and unmatched coding skills, offering PhD-level expertise across all subjects.

OpenAI has officially launched its next-generation AI system — ChatGPT 5, also known as GPT-5 — marking what CEO Sam Altman calls a revolutionary leap in artificial intelligence capabilities. The announcement, made late Thursday, confirms that GPT-5 will replace all earlier OpenAI models, including the o series and previous ChatGPT versions, bringing a unified, more powerful AI experience to users worldwide.

In Altman’s words, GPT-5 represents the first time it truly feels “like you’re talking to an expert in any topic.” Drawing a vivid comparison, he said, “It reminds me of when the iPhone went from those giant, old pixels to the Retina Display.”

A Unified System for All Users

According to OpenAI, GPT-5 integrates all AI capabilities into a single system, featuring:

  • A smart, efficient model for most questions.
  • A deeper reasoning mode (GPT-5 Thinking) for more complex problems.
  • A real-time router that automatically decides which mode to use, based on the nature and complexity of the conversation.

This design eliminates the need for separate toggles for deeper reasoning, which was previously limited to premium modes in earlier ChatGPT versions.

PhD-Level Knowledge, Everywhere

OpenAI claims GPT-5 delivers PhD-level expertise across all subjects, from science and mathematics to literature, history, and art. The company emphasizes that the model not only provides more accurate answers but also significantly reduces hallucinations — AI-generated factual errors.

When tested with web search on anonymized prompts from real ChatGPT traffic, GPT-5’s standard mode was 45% less likely to produce a factual error compared to GPT-4o. In Thinking mode, the improvement soared to 80% fewer errors than OpenAI’s o3 model.

Coding Powerhouse

One of GPT-5’s standout improvements is in software development. OpenAI describes it as “our strongest coding model to date,” capable of building sophisticated, responsive, and aesthetically appealing websites, apps, and games from a single prompt. The system also excels at debugging large codebases and creating polished frontend designs — a leap forward for developers.

Availability and Pricing

ChatGPT 5 will roll out to all users — including those on the free plan — starting today. The system comes in three tiers:

  1. ChatGPT 5 Regular – Default access for both free and paid users.
  2. ChatGPT 5 Mini – Automatically enabled when a user’s Regular tier usage is exhausted.
  3. ChatGPT 5 Pro – Available for a $200/month subscription, offering the highest performance tier and the ability to force “Thinking Mode” at will.

With GPT-5 as the new default, earlier models like GPT-4o, GPT-4.1, GPT-4.5, o3, and o4-mini will be retired for signed-in users.

By combining advanced reasoning, dramatically reduced factual errors, and unmatched coding capabilities, GPT-5 sets a new standard for conversational AI — one that might just feel like chatting with a world-class expert.

Next Story
Share it