I recently made the switch from GPT4 Mini to GPT5 Mini for my AI agent, and the results have been nothing short of amazing. The new model has enabled my agent to handle complex tasks with ease, including answering customer emails and saving them as drafts. The accuracy and detail of the responses are impressive, and I’m excited to see how this upgrade will improve my workflow.
One of the biggest challenges I faced with GPT4 Mini was its inability to handle source information well. This limited its ability to provide accurate answers, and I had almost given up on the project. But with GPT5 Mini, I’ve seen a significant improvement in performance. The model is able to understand and process complex information, making it a valuable tool for my business.
Of course, there are some trade-offs to consider. The process of setting up the new model required me to upload a government ID and take phone pictures of my face from different angles. While this was a bit unsettling, I understand the need for security measures.
As I continue to refine my agent, I’m excited to explore the possibilities of what it can do. I’m considering adding more sub-agents to handle specific tasks, such as managing rental calendars. The potential for automation and increased efficiency is vast, and I’m looking forward to seeing how this upgrade will impact my business in the long run.