At its recent I/O developer conference, Alphabet (NASDAQ: GOOGL) announced a series of upgrades to its AI creation tools aimed at developers. The updates are designed to lower the barrier to entry and improve efficiency for generating multimedia content through its latest Gemini family of models.
A central part of this update is the launch of the Gemini 3.5 Flash model. This model is now integrated as the default within the Gemini application and the AI mode of Google Search. It is optimized for coding, agent-based tasks, and real-world workflows, with a notably faster output speed compared to its predecessor. The company states it is up to four times faster than other leading models. Developers can use it to quickly generate web interfaces, interactive components, and even transform complex research papers into visual learning cards, offering a more balanced model choice for scaling AI applications.
In the realm of video and multimodal creation, the company introduced the new Gemini Omni model. This model can understand and process inputs of text, images, audio, and video to generate coherent video content. Its most prominent feature is support for conversational editing, where users can simply describe their desired changes—such as replacing a character, adjusting lighting, or altering a scene—in natural language, and the model will automatically perform the edits. To address safety and provenance challenges, all videos generated by Omni will include an invisible SynthID digital watermark to help identify AI-generated content. Currently, the Gemini Omni Flash version is available to subscribers and will be offered to developers via API in the future.
Furthermore, the company is integrating its AI capabilities deeper into development and design workflows. For example, Google Labs has released a free design tool called Stitch, which allows developers to generate user interface prototypes through natural language conversations. These prototypes can then be exported with a single click to Figma or as HTML/CSS code, significantly shortening the process of turning design concepts into usable assets.
Comments