Guys what is the status of these models performance as they are way too slow.
Their default speed should match GPT 5.6 to feel natural
Hey Andrew, thanks for the feedback. We are always striving to improve the performance, speed, and feel of our models. We appreciate your feedback, and that helps us improve. In the meantime,
- Enable Fast in the model picker for priority serving and faster output at a higher usage rate.
- Start a new chat when conversations become long, since larger contexts can increase the wait.
- If Grok repeatedly checks frontend work in the browser, add a rule such as: “Do not verify changes in the browser or take screenshots unless I explicitly ask.”
If a run stays on “Taking longer than expected” for over a minute, please feel free to share its full Request ID and model here so I can investigate it.