top of page
  • Voltaire Staff

Google's Bard to allow text-to-image generation


Google's AI-powered assistant Bard will now allow generation of images, complete with a watermark in their pixels to tell them apart from real artwork.


The change came as a part of a series of updates, including one that will allow Bard to offer support in over 40 languages.


"Today Gemini Pro in Bard will be available in over 40 languages and more than 230 countries and territories, so more people can collaborate with this faster, more capable version of Bard," Google said in a statement on Thursday.


The languages will include various Indian languages such as Hindi, Bengali, Gujarati, Kannada, Malayalam, Marathi, Tamil, Telugu, and more.


Users can now instruct Bard to generate images using Google's Imagen 2 text-to-image model. Despite initial expectations of Bard being powered by the more advanced Gemini Ultra model, it is currently fueled by Google's Gemini Pro large language model, with the latter still undergoing development.


"Consistent with our AI Principles, image generation was designed with responsibility in mind. For instance, to ensure there’s a clear distinction between visuals created with Bard and original human artwork, Bard uses SynthID to embed digitally identifiable watermarks into the pixels of generated images," the tech giant said.


Unlike some other AI image generation tools, Bard's service is free, making creative exploration accessible to a wider audience.


Google said it is also expanding Bard's double-check feature, allowing evaluation of responses across the web, now available in more languages.


Users can click on the Google icon and the highlighted phrase to explore additional supporting and contradicting arguments from the web.


In addition to Bard, Google is introducing generative AI features on Google Maps, enabling the curation of location lists based on input prompts and offering place recommendations upon request.


Currently in the experimental phase, these features are exclusively available in the US, with plans for a broader rollout in the upcoming months.


Google has positioned Bard as a strong contender against OpenAI's ChatGPT Plus, recognised for its ability to run GPT-4 and facilitate image generation through DALL-E 3 integration.


While both chatbots have performed well, Bard's earlier absence of text-to-image features slightly favoured ChatGPT Plus.

bottom of page