Gemini CLI tutorial and Gemini web UI comparison

This beginner walkthrough introduces Gemini CLI as a terminal coding assistant and compares it with Gemini's web interface. The demonstration starts in Windows PowerShell with the CLI already installed and the presenter signed in through Google authentication. It does not walk through installation or credential setup.

The presenter explains natural language chat, shell mode triggered with an exclamation mark, and planning through /plan. The command tour covers /help, file references with @, and /init for project analysis and a Gemini.md context file. It also covers model selection, memory commands, copying the last response and quitting with a session summary. The model menu shown includes Gemini 3 and Gemini 2.5.

The comparison uses prompts for the current date and gold prices in New Delhi, a pizza image, and an HTML page whose button changes the background color. The presenter reports faster web UI responses for the search task. In the image example, the web UI produces an image while the CLI supplies a suggested prompt.

The HTML example shows the practical difference in file handling: after permission approval, the CLI creates bgchanger.html in the local project directory. The web interface supplies code and then a downloadable file. These examples concern terminal access and local file changes; they do not demonstrate running a local LLM or offline inference.