Google Gemini app update: Better AI videos thanks to multiple reference images

While OpenAI Sora 2 is still not available in Europe, Google is working on improving the performance of its competing model called Veo. After Veo 3.1 launched in mid-October, the model has now also been integrated into the Gemini app for mobile and desktop devices. With the latest update, Google is addressing one of the biggest challenges and gives users more precise control. As Google announced on its official X account for the Gemini app, users can now upload multiple reference images in addition to their text prompts.

For example, one image can be used as a reference for the character, another for the background and a third for the overall visual style. The text prompt can then give more precise instructions on how these elements should be combined in the AI-generated video.

Familiar AI feature now available to more users

This feature is not entirely new, since the "Ingredients to Video" feature was already integrated into Google’s AI filmmaking tool Flow back in October. After the latest update, this feature is no longer limited to Google’s specialized developer tools like Flow or Vertex AI, but is now built directly into the regular Gemini app.

According to Google, the rollout of this new feature has already started.