As Google promised in August, the artificial intelligence model for generating images has been made available to all users of the Gemini service.
Image source: x.com/GeminiApp
The updated Imagen 3 model, presented in May at the Google I/O 2024 event, according to the developer, is distinguished by three aspects:
- Stunning image quality – sharp details, vibrant colors and reduced imperfections;
- Realistic visual effects – bright photorealistic details;
- A variety of styles – from classic oil to modern digital art.
To create images, users need to use words like “draw” or “create” in queries to Gemini; you can specify the desired style, for example, “photorealistic watercolor” or “cartoon illustration”; Detailed explanations will also help. In response, the chatbot will explain that the picture is created using Imagen 3; In addition, the downloaded file will be marked SynthID. Users of free accounts can also use the image generator, but the options for them are still limited – the system will refuse to create pictures with people. English-speaking users of Gemini Advanced, Business and Enterprise subscriptions were the first to gain access to Imagen 3.