August 11, 2025

OpenAI has officially launched GPT-5, the latest evolution of its flagship language model—and it’s already making waves. CEO Sam Altman calls it “a significant step along the path to AGI,” though he’s careful to note that it doesn’t quite reach artificial general intelligence. Still, GPT-5 is smarter, faster, and more capable than any model before it.
So what’s new?
Smarter, Faster, More Human
GPT-5 feels like a leap forward in conversational AI. Altman likened the jump from GPT-4 to GPT-5 to the iPhone’s upgrade to Retina display—suddenly, everything feels sharper. He described it as “talking to an expert in any topic, like a PhD-level expert.”
The model boasts a 256,000-token context window, meaning it can handle longer conversations, documents, and code without losing track. It also hallucinates less, thanks to improved safety mechanisms and better instruction-following.
Meet the GPT-5 Family
OpenAI didn’t just release one model—they launched a whole lineup:
• GPT-5: The full-featured flagship model.
• GPT-5-mini: A lightweight version for faster, simpler tasks.
• GPT-5-nano: The fastest and cheapest option, available only via API.
• GPT-5-pro: A more powerful version for Pro-tier users.
• GPT-5-thinking: Designed for longer, more complex queries.
Most users won’t need to choose between models—the ChatGPT interface now automatically routes queries to the right version based on complexity and subscription tier.
New Features for Users
OpenAI is rolling out several fun and functional updates:
• Chat Personalities: Choose from Cynic, Robot, Listener, and Nerd (coming to Advanced Voice Mode).
• Customization: Pick your chat color.
• Google Integration: Pro users can connect Gmail, Contacts, and Calendar starting next week. Other tiers will follow.
ChatGPT will automatically reference these tools when relevant—no need to manually select them.
Adoption Statistics
OpenAI’s reach continues to grow:
• 700 million weekly active users
• 5 million paying business users
• 4 million developers using the API
Final Thoughts
Nick Turley, the head of ChatGPT, summed it up like this:
“The vibes of this model are really good, and I think that people are really going to feel that – especially average people who haven’t been spending their time thinking about models.”
GPT-5 isn’t just a technical upgrade, it’s a shift in how we interact with AI. Whether you’re a developer, a business user, or just curious, this release marks a new chapter in human-AI collaboration.
Initial user reactions to GPT-5 have been mixed, with both excitement and frustration surfacing across online communities. Here’s what people are saying:
Positive Impressions
• Expert-Level Responses: Some early testers and reviewers have praised GPT-5 for feeling like “talking to a PhD-level expert,” especially in technical and academic domains.
• Improved Safety and Accuracy: Users have noted fewer hallucinations and better instruction-following, particularly in coding and health-related queries.
• Expanded Capabilities: The new 256,000-token context window has been well received for handling longer conversations and documents.
User Backlash
Despite the upgrades, many users have voiced strong dissatisfaction:
• Shorter, Less Useful Replies: A popular Reddit thread titled “GPT-5 is horrible” has thousands of upvotes and comments criticizing the model for overly brief answers and reduced personality.
• Usage Limits: Plus users are frustrated by stricter message caps (e.g., 200 messages per week), which feel more restrictive than previous versions.
• Loss of Model Choice: Users lament the inability to switch back to older models like GPT-4 or GPT-4o, which some preferred for their tone and flexibility.
• “Feels Like a Downgrade”: One user summed up the sentiment by saying, “Answers are shorter and not any better than previous models… it feels like a downgrade branded as the new hotness”.
GPT-5 is free to use for all users, including those currently on the free tier. However, free users have usage limitations, such as a limited number of messages per hour and one “Thinking” message per day, before switching to a smaller model (GPT-5 mini). Paid plans (Plus, Team, Enterprise) offer higher usage caps and access to more advanced features.
Let’s wait a few weeks and see what Google’s Gemini comes back with. The AI battle has only just begun.
Resources:
https://www.wired.com/story/openais-gpt-5-is-here/
https://www.tomsguide.com/ai/chatgpt-5-is-here-openais-most-powerful-ai-model-just-changed-everything

Sidebar: Back to the “GPT Family”
Pam didn’t know what API is, so she asked GPT, and it told her:
When you see “ChatGPT-5 nano is available only via API”, it means:
• You can’t use that model directly inside the ChatGPT web app.
• You can use it if you connect to OpenAI’s API from your own code or tools.
• The API endpoint would let you send text (or other input) to the model and receive responses, just like ChatGPT, but you control it from your own program, script, or app.
Think of it like: ChatGPT here is the pre-built car you can drive immediately, but the API is the engine you can install in your own vehicle.

David Snell joins Rob Hakala and Beth Foster of the South Shore’s Morning News on 95.9 WATD fm every Tuesday at 8:11
You can listen to this broadcast here: https://actsmartit.com/gpt-5-most-powerful-yet/