Google is expanding Gemini’s experimental features with some new modes. If we go by the details spotted by Testing Catalog. The updates go beyond the previously known “gem pix” image editing option connected to the upcoming nano-banana model.
The highlight of the addition is the new Agent Mode, which was first teased at Google I/O. The latest description confirms that it will support autonomous exploration, planning, and execution. This brings Gemini closer to the agent-like workflows seen in its rivals like ChatGPT.
