Mixing GPT-6 Astra and Opus 5.5 gives better results
The poster advises against taking sides between models and suggests orchestrating GPT-6 Astra with Opus 5.5, saying a friend's work looks like it came straight out of a movie.
Despite all the OpenAI vs Anthropic fighting, when you put GPT-6 Astra and Opus 5.5 together, the results can look straight out of a movie. I’m looking at some of my friends’ work right now and it’s honestly on another level. They know how to orchestrate teams of AIs really well. So my advice to everyone is simple: don’t get biased against models. Use all of them, figure out their strengths and weaknesses for yourself, and if you can, mix them together and make them work as a team.