At Zide, we designed AI model choice to belong right inside your conversations

We all know that there is no single best model for every development task. We make choices every day as to how and where to start a project or to continue with a specific task or issue.

Sometimes we want a fast answer to a small question. Other times, we need a model to spend more time understanding a large repository. I often prefer one model for planning, one for implementation, one for issues, one for code changes, and another for reviewing the results. The cost, the speed, context size, and the number of tools used all matter for a developer differently depending on the exact work at hand.

In many of the AI development apps available, the model choice is typically treated as an application setting that you configure away from the conversation. Zide has always made sure to keep this choice easier to make and closer to the task.

One specific feature we developed that makes this even simpler for the user is our /model command for Zide Assist. It opens a searchable, keyboard-navigatable model picker directly from the conversation prompt box.

I often find that AI coding tools will hide model selection behind labels such as fast, smart, or deep reasoning. Those shortcuts can absolutely be useful for developers but here at Zide, we prefer to make the choice clear and explicit. You can start work with a quick, economical model, move to a more capable one when the task demands it, and make that decision anytime and within the context of the work open in front of you. This approach, we feel, makes the model not define your workspace. It is one part of your workflow, and it should be easily changed as your tasks change.

With Zide, we provide the assistant, the repository, the tools, controllable context, and all the code review surfaces. The model you choose is the AI for the work in front of you.