How will GPT-6 Sol and Astra split thinking effort levels?
The post asks whether OpenAI will build a more granular thinking-effort ladder for Sol and Astra, or keep Plus users permanently capped at Sol.
GPT-6 Sol is coming soon, and I’m really curious how OpenAI plans to balance the different Thinking Effort levels between 6 Sol and 6 Astra in ChatGPT Chat. 🤔 Will they build a more granular ladder in Chat and let users choose based on compute and usage limits? Or will they simply keep the current structure, with Astra based 6 Pro reserved for Pro users while Plus users are permanently capped at 6 Sol? This product decision is going to matter a lot. 👀
This post quotes another post: “GPT-6 Sol internal test: zero-shot output in 15 minutes and 60k tokens”