Google Bard gets Gemini Pro update, image-generation capabilities

Google’s Bard chatbot is all set for another update. The company announced today that Bard will now be powered by the Gemini Pro models as well as the Imagen 2 model, the latter of which will now allow it to generate images in addition to text. Here’s all you need to know.

Google Bard: New features to look out for

Google Bard is now going to be powered by the Gemini Pro Large Language Model (LLM). It will now support over 40 languages, including Hindi, Japanese, Arabic, Chinese, Portuguese, Spanish, Tamil Telugu, and Malayalam. The new languages supported on the chatbot, in addition to text generation, will also work with the double-check feature.

The double-check feature, for those unaware, allows you to evaluate if the results generated by Google Bard for your queries are accurate, using Google Search side-by-side.

ALSO READ: Google Bard now available in Hindi and other Indian languages: Here’s how to try it out

Now, Google didn’t exactly mention any new capabilities that the chatbot will now have explicitly, however, it did mention that it will be better in terms of summarising content, brainstorming, as well as writing.

In addition to this, Google Bard will now also get image generation support via the Imagen 2 model, which was also released in December. The feature will work the way Midjourney or Open AI’s Dall-E 3 model works, wherein you can give it written prompts and it will generate an image based on the same.

Images created using this feature will have a SynthID digital watermark on them developed by Google’s DeepMind labs, embedded in the pixels. This is to ensure that the images aren’t manipulated or misused such that they may cause any harm or offense to someone.

ALSO READ: Google Bard explained: All about the new ChatGPT rival

At the time Google Bard was originally launched in February last year, it was powered by Google’s LaMDA (Language Model for Dialogue Applications). Later in the year, it received another update and was powered by a new LLM called PaLM 2. Gemini Pro is the third LLM to power Google Bard, and the same will be made available in over 230 countries.

What are the Gemini Large Language Models?

Google announced its Gemini LLMs in December 2023. The line-up consists of three LLMS – the flagship Gemini Ultra, the Gemini Pro, and Gemini Nano.

Gemini Nano is designed to run on small devices, such as smartphones including the likes of the Google Pixel 8 series, while the Gemini Pro is designed more for cloud-based generative AI platforms (like Google Bard). Gemini Ultra, on the other hand, is still currently undergoing testing.