- My goal: See which models are called by which Agents, Apps and workflows so I can find expensive calls and swap them out for cheaper ones
- Issue: In my OpenAI analytics, I can see something started using our Retool key to call gpt-4 rather than gpt4o-mini, which is much more expensive, blew our budget and now has taught us the lesson we need to whitelist models. We cannot easily see what in Retool is using 4 and not 4o-mini, our default. We have whitelisted models and are now waiting for something in Retool to break so we can find it.
- Steps I've taken to troubleshoot: None, not an issue, feature request
- Additional info: (Cloud or Self-hosted, Screenshots)
