Some big changes have just landed for GPT-5 and the model picker in ChatGPT:
⚙️ New Speed & Thinking Controls
You can now choose between:
- Auto – ChatGPT decides for you (best for most people)
- Fast – quicker responses, less deep thinking
- Thinking – longer reasoning time for more complex tasks
📈 Thinking Mode Limits
- Plus users now get 3,000 messages/week with GPT-5 Thinking
- After that, you’ll still have extra capacity with GPT-5 Thinking mini
- GPT-5 Thinking has a 196k token context limit (great for bigger projects)
- OpenAI may adjust rate limits in future depending on usage
🔄 Model Picker Changes
- GPT-4o is back by default for all paid users
- Paid users now have a “Show additional models” toggle in ChatGPT web settings—turn this on to get access to models like o3, o4-mini, 4.1, and GPT-5 Thinking mini
- GPT-4.5 remains Pro-only due to GPU constraints
Have you tried switching between Fast and Thinking yet?
What’s your first impression?