I’m done with ChatGPT and OpenAI. As a heavy user, I’ve relied on their models for various tasks, from work to personal projects. But with the release of GPT-5, I’ve been left feeling disappointed and frustrated. The new model is unusable for me, providing garbage responses that are far from the quality I’ve come to expect.
My workflow was perfect, with specific instruction templates for different tasks. I knew exactly which model to pick for which task, and it worked beautifully. But GPT-5 feels like a cheap, restrictive router, and the API’s dirt-cheap pricing screams cost-saving move.
I’ve had to move my entire workflow over to Gemini 2.5 Pro, and I’m absolutely impressed. It’s an absolute powerhouse, handling every task I threw at it brilliantly. The huge context window, which lets me feed it a ton of info that it actually uses, is a game-changer. And the fact that it pushes back when I go outside the boundaries of the context I provide is a great feature for academic work.
I know this post might sound like a Gemini ad, but it’s only because I’m genuinely so pleased with the alternative after being so let down. If you’re feeling burned by the recent OpenAI changes, I seriously suggest you look around. You might be surprised at what you find.