Yes, the best approach is to let the top models audit/orchestrate the agents rather than have them to doing the work directly. Personally, I love opus 4.6s output, fable was too annoying with compliances and legal constraints.
I found GPT-6 Astra to be a better orchestrator than any other model on the market right now. But only that. Otherwise it burns my $100 subscription like it's nothing.
Took myself out of Claude-slop world and not sure when I will be going back
Yes, the best approach is to let the top models audit/orchestrate the agents rather than have them to doing the work directly. Personally, I love opus 4.6s output, fable was too annoying with compliances and legal constraints.
Yes precisely
I found GPT-6 Astra to be a better orchestrator than any other model on the market right now. But only that. Otherwise it burns my $100 subscription like it's nothing.
Took myself out of Claude-slop world and not sure when I will be going back
Claude works really well for me tbh, really happy with it. Maybe the outputs are industry dependent as well.