Nice looking app, downloaded and checking it out. Fyi there's someone else on github keeping track of some LLM benchmarks too:

https://github.com/ggml-org/llama.cpp/discussions/4167