If you’re curious, you can check the progress of many open source LLM’s and how they perform on various evals here:

https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderb...