The AI landscape is changing so rapidly that LLMs and Providers are constantly leapfrogging each other on an almost monthly basis. So what we recommend at AssetBytes is to architect your code and AI platforms in such a way that you can easily swap out LLMs without having to re-write your application.
There are a few AI Leaderboards available on the internet which try to rank the LLMs but not all of them are reliable or accurate or updated frequently. They key factor to keep in mind when looking at these leaderboards is that which model you choose is dependent on your specific use case or domain and take into account 3 main factors
- Model Intelligence or Accuracy/Reliability
- Model Speed
- Model Cost
The good folk at Artificial Intelligence have leaderboards which take into account models for many common modalities and rank/filter them according to the above factors as below