I just added a quick training showing you a simple way to make sure each of your AI agents is using the right model for the job. New models are constantly being released, but that does not mean you should automatically switch every agent to the newest one. Different agents do different work. Your Executive Assistant may need stronger reasoning. A research agent may need something different. A scheduling or repurposing agent may not need the most powerful model at all. So instead of guessing, I had my Executive Assistant review every agent I currently use and recommend: - the best model for each agent - the right reasoning level - why that setup fits the job - what I should update You can do the same thing inside Codex, ChatGPT, Claude, or whatever platform you are currently using. Use a prompt like this: “Review each of the AI agents I currently use. Based on the purpose of each agent, the tasks it completes, the tools it uses, and the output I need, recommend the best available model and reasoning level for each one. Explain why and flag anything I should update.” Then give it a quick description of each agent. For example: - Executive Assistant - Content Brain Dump Agent - Research Agent - Repurposing Agent - Scheduling Agent - Social Media Audit Agent It will give you a simple audit of your current setup. The bigger lesson is: Match the model to the job instead of assuming every agent needs the same setup. This is a small adjustment, but it can help your agents work faster, produce better results, and avoid wasting unnecessary usage. Watch the Loom, run the audit, and then comment below with: - which agents you reviewed - what model changes it recommended - what you updated Don’t just watch the training. Go audit your setup and get the quick win.