For me good enough is usually good enough (and pretty excellent). Unless I ask for something really complex, then I use perplexity in research mode. But I have

For me good enough is usually good enough (and pretty excellent). Unless I ask for something really complex, then I use perplexity in research mode. But I have everything in OpenWebUI so I can choose with a touch of a button. Sometimes I wonder whether GPT could have done better and I try it and it's usually not significantly better than llama 3.1 8b