Google Gemini 3.0 and GPT 5.1

Early Gemini 3.0 Reviews Are Positive. GPT 5.1 Is A Non-Event

Early Gemini 3.0 reviews are largely positive. I have yet to try it as I do not receive early releases. I have kicked the tires on GPT 5.1, which was released yesterday. I have not noticed a difference between GPT 5.1 and 5.0, although 5.1 is supposed to process more quickly.

I asked 5.1 to clone Glassdoor as an autonomous coding exercise. As you can see from the below, GPT 5.1 is good at writing front-end html. GPT churned out the html in approximately 3 minutes.

While the Website looks great, none of the buttons are operational, as the html does not link to real databases, real APIs, or anything of the sort.

GPT 5.1 could help you write the back-end code and middleware jobs if you had the required API keys and commercial partnerships. However, developers would need to supervise the LLM to validate its output.

I would never allow any of the models to write large chunks of code and drop them into a production environment. Everything the models churn out must first be validated.

---