GPT-5 is live and it is the best model that 99 percent of the world has ever used 🚀🚀 We've been doing hands-on alpha testing internally @every for the last few weeks. We've used it for everything from coding to writing. Here's the only guide you know for what it's good at (and what it's not good at): - It's fast, really fast. It's also simple and powerful enough to be a daily driver. - OpenAI got rid of the model picker in ChatGPT, now all you see is GPT-5, and you won't want to go back. - It's a hybrid chat and reasoning model and it is available for free today to everyone using ChatGPT. This will be the first time that millions of people use a reasoning model—a gigantic moment. - In ChatGPT, GPT-5 returns comprehensive and readable answers with logical subsections, judicious use of white space, and bolded text to help you find what you need quickly. (It will remind you of answers you get from o3, but without the obsession with tables.) - In Canvas, GPT-5 can quickly one-shot front-end apps, making it the introduction of vibe coding to millions of people who would never pay for a Claude subscription or try coding app Loveable. - In the API, GPT-5 is priced to make rivals sweat. GPT-5-mini undercuts Google's Gemini 2.5 Flash, and GPT-5 Standard comes in at 1/12th the cost of Claude 4 Opus. - But if you're vibe coding apps with four Claude Code instances running at once, using GPT-5 in its current form is not going to become your daily driver for frontier AI coding. It's more like a skilled utility player than your new MVP. - GPT-5 is a very good programmer. It's incredibly useful as a pair programmer, especially in AI-powered integrated development environments (IDEs) like Cursor. It's great for engineers from traditional backgrounds who want an AI to help collaborate on code. And in our testing it excels at research and debugging complex issues. - But GPT-5 isn't as good as Opus 4.1 as a true programming agent. It is more cautious than Opus and isn't as comfortable working independently for long periods in our testing. - Both Cursor and OpenAI's command line interface tool Codex CLI are not on the same level as Claude Code. Both were built for programmer-AI pair programming, not true delegation. The model is extremely smart, just not yet built for this use case. Overall, GPT-5 is a huge leap forward for consumers and for developers in the API. But as a true agentic coding tool it needs some work.