Google is enhancing its Bard chatbot with the capability to generate images, aligning its features with those of its competitor, ChatGPT Plus, which has offered this functionality for several months.

Bard users will now be able to create images through prompts, utilizing Google’s Imagen 2 text-to-image model. Google says that Imagen 2 has been “designed to balance quality and speed, delivering high-quality, photorealistic outputs.”

Although Bard has been upgraded to the Gemini Pro large language model, there was an initial assumption that image generation would be powered by the more advanced Gemini Ultra model, which is still under development.

Image created with Google Bard

Previously, Gemini Pro could only be used with Bard with support for English, but now it will be available in all languages and regions where Bard is available. Google says that the upgrade to Gemini Pro gives Bard “more advanced understanding, reasoning, summarizing and coding abilities”, as per the company’s blog post.

The best part is that image generation features in the new Bard are absolutely free, unlike ChatGPT Plus. This is similar to Microsoft’s Copilot, which offers the capabilities of GPT 4, including image generation, for free.

Using Bard’s AI image generation is just like any other chatbot; users simply have to enter a prompt to describe their idea and Bard will bring it to life. To ensure the origin of these images is traceable, Bard incorporates SynthID technology, embedding digital watermarks within the image pixels, providing a clear method to identify them as AI-generated.

GIF of a prompt that says "create an image of a dog riding a surfboard." Image results appear.

Google has introduced a feature allowing users to verify Bard’s responses against Google search findings in over 40 languages by clicking the G icon. Upon activation, Bard assesses the web for content that corroborates its answers. Users will see highlighted text, enabling them to delve deeper into information that either supports or refutes the findings sourced from Google search.

Other than feature upgrades, Bard is also expanding to more countries and languages. Although image generation is currently limited to English, the chatbot’s services have been extended to support over 40 languages across more than 230 countries. Previously, the chatbot functioned solely in English and was accessible in 170 countries. Among the languages it now supports are Arabic, Bengali, Tamil, and Urdu.