AI chatbots are more popular than ever, and there are plenty of solid options out there to choose from beyond OpenAI’s ChatGPT. One particularly strong competitor is Google’s Gemini AI, which used to be called Google Bard. This AI chatbot pulls information from the internet and runs off the latest Gemini language model created by Google.
What is Google Gemini?
Bard, or Gemini as the company now calls it, is Google’s answer to ChatGPT. It’s an AI chatbot designed to respond to various queries and tasks, all while being plugged into Google’s search engine and receiving frequent updates. Like most other chatbots, including ChatGPT, Gemini can answer math problems and help with writing articles and documents, as well as with most other tasks you would expect a generative AI bot to do.
What happened to Google Bard?
Credit: Google / Joshua Hawkins
Nothing happened—Google just changed the name. Bard is now Gemini, and Gemini is Google’s home for all things AI. The company says it wanted to bring everything into one easy-to-follow ecosystem, which is why it felt the name change was important. You can still access Gemini through the old bard.google.com system, but it will now redirect you to gemini.google.com.
How does Gemini work?
Much like ChatGPT, Gemini is powered by a large language model (LLM) and is designed to respond with reasonable and human-like answers to your queries and requests. Previously, Gemini used Google’s PaLM 2 language model, but Google has since released an update that adds Gemini 1.5 Flash and Gemini 1.5 Pro models, the search giant’s most complex and capable language models yet. Running Gemini with multiple language models has allowed Google to see the bot in action in several different ways. Gemini can be accessed on any device by visiting the chatbot’s website, just like ChatGPT, and is also available on Android and iPhones via the Gemini app.
Recommended AI courses:
Who can access Google Gemini?
Gemini is currently available to the general public. Google is still working on the AI chatbot, and hopes to continue improving it. As such, any responses, queries, or tasks submitted to Gemini can be reviewed by Google engineers to help the AI learn more from the questions that you’re asking.
To start using Gemini, simply head over to gemini.google.com and sign in. Users who subscribe to Gemini Advanced can utilize the newest and most powerful versions of the AI language model. (More on that later.)
What languages does Gemini support?
Gemini 1.0 Pro currently supports over 40 languages. Google hasn’t said yet if it plans to add more language support to the chatbot, but a Google support doc notes that it currently supports: Arabic, Bengali, Bulgarian, Chinese (Simplified / Traditional), Croatian, Czech, Danish, Dutch, English, Estonian, Farsi, Finnish, French, German, Greek, Gujarati, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Kannada, Korean, Latvian, Lithuanian, Malayalam, Marathi, Norwegian, Polish, Portuguese, Romanian, Russian, Serbian, Slovak, Slovenian, Spanish, Swahili, Swedish, Tamil, Telugu, Thai, Turkish, Ukrainian, Urdu, and Vietnamese.
Gemini 1.5 Pro supports 35 languages and is available in over 150 different countries and territories. The supported languages include Arabic, Bulgarian, Chinese (Simplified / Traditional), Croatian, Czech, Danish, Dutch, English, Estonian, Farsi, Finnish, French, German, Greek, Hebrew, Hungarian, Indonesian, Italian, Japanese, Korean, Latvian, Lithuanian, Norwegian, Polish, Portuguese, Romanian, Russian, Serbian, Slovak, Slovenian, Spanish, Swahili, Swedish, Thai, Turkish, Ukrainian, and Vietnamese.
(Note: At the time of this article’s writing, Google Gemini Advanced is only optimized for English. However, Google says it should still work with any languages Gemini supports.)
What features does Gemini offer?
Like ChatGPT, Gemini can answer basic questions, help with coding, and solve complex mathematic equations. Additionally, Google added support for multimodal search in July, allowing users to input pictures as well as text into conversations. This, along with the chatbot’s other capabilities, enables it to complete reverse image searches. Google can also include images in its answers, which are pulled from the search giant’s online results.
Google also previously added the ability to generate images in Gemini using its Imagen model. You can take advantage of this new feature by telling the bot to “create an image.” This makes the chatbot more competitive with OpenAI, which also offers image generation through DALL-E.
During Google I/O 2024, Google also showed off plans to expand that multimodal support for Gemini to include video and voice, allowing you to chat with the AI chatbot in real-time, similar to what we’re already seeing with ChatGPT’s new GPT-4o model.
Is Gemini connected to the internet?
Yes, Google Gemini is connected to the internet and is trained on the latest and most up-to-date information found online. This is obviously a nice advantage over ChatGPT, which just added full access to the internet back in September, and only for paid users who subscribe to its GPT-4 model.
How accurate is Google Gemini?
Now that the chatbot is using Gemini 1.0 Pro and Gemini 1.5 Pro, it’s expected to be one of the most accurate chatbots available on the web right now. However, past experiences with Gemini have shown that the bot is likely to hallucinate or take credit for information that it found via Google searches. This is a problem that Google has been working to fix, and the company has managed to improve the results and how they are handled.
However, like any chatbot, Gemini is still capable of creating information that is untrue or plagiarized. As such, it is always recommended you double-check any information that chatbots like Gemini provide, to ensure it is original and accurate.
Is Gemini free to use?
Gemini is currently free to use, but Google also offers a subscription-based plan that allows you to take advantage of its best AI yet, Gemini Advanced. The service is available as part of Google’s new Google One AI Premium Plan, which currently runs for $19.99 a month, putting it on par with ChatGPT Plus. The advantage here, of course, is that you also get access to 2TB of storage in Google Drive, as well as access to Gemini in Gmail, Docs, Slides, Sheets, and more. This feature was previously known as Duet AI, but it has also been rounded up under the Gemini umbrella.
There’s an app for that
Google also launched a dedicated Gemini mobile app for Android. iPhone users can access Gemini through the Google app on iOS. Currently, the Gemini mobile app is only available on select devices and only supports English in the U.S. However, Google plans to extend the available countries and languages the Gemini app supports in the future. Additionally, the mobile app supports many of the same functions as Google Assistant, and Google is positioning it to replace Assistant with Gemini in near future.
How does Gemini compare to ChatGPT?
Gemini is a solid competitor for ChatGPT, especially now that Gemini should return results more akin to those in GPT-4. The interface is very similar, and the functionality offered by both chatbots should handle most of the queries and tasks that you throw at either of them.
Even with Google’s paid plan, Gemini is still a more accessible option, as its free models are more similar to GPT-4 than ChatGPT’s free option is. That said, OpenAI is starting to roll out a version of GPT-4o to all users, even free ones, but it will have usage limits and isn’t widely available yet.
For now, Gemini presents the fewest barriers to internet access, and can use Google as a search engine. When ChatGPT does connect to the internet, it utilizes Bing as a search engine instead of Google.
Google did share some information about how Gemini compares to GPT-4V, one of the latest versions of GPT-4, and said it actually achieves more accurate results in several fields. But as no trustworthy tests are yet available for how Gemini 1.5 Pro compares to GPT-4o, it’s unclear exactly how the two newest models from Google and OpenAI compare when placed head to head. Google Gemini 1.5 Pro does offer a maximum context-token count of one million, so it can handle much longer context documents than ChatGPT can now. And Google isn’t stopping there, as it plans to offer a Gemini version with support for two million context tokens, which it is already testing with developers.
Ultimately, it’s hard to say exactly which one is better, as they both have their strengths. I’d recommend trying to complete whatever task you want to accomplish in both, and then seeing which one works best for your needs. Also, keep in mind that some of the most impressive features that Gemini and ChatGPT offer are not fully available yet. For its part, Google is working on other AI-driven systems, which it could possibly include in Gemini down the line. These include MusicLM, which uses AI to generate music, something the tech giant showed off during Google I/O 2024.
