Two months after bringing ChatGPT to iOS, OpenAI LP today launched an Android version of its artificial intelligence assistant.

The Android app is currently accessible for users in the U.S, India, Bangladesh and Brazil. OpenAI will extend availability to additional countries over the next week. The iOS version was available for download in more than 150 countries as of late May.

Today’s launch of ChatGPT on Android was not unexpected. OpenAI first disclosed plans to add support for the operating system after launching its iOS app. Last week, the startup created a listing in Google LLC’s Play Store that allowed users to preregister for the Android client.

Both the iOS and Android versions of ChatGPT are available at no charge. Consumers can, however, use the apps with a paid ChatGPT Plus subscription. The subscription is priced at $20 per month and offers faster chatbot response times along with priority access to updates.

OpenAI’s new Android app offers a similar set of features as its iOS counterpart. Users can sync their chat histories across multiple devices and interact with ChatGPT using voice commands. The latter feature is powered by Whisper, an open-source speech recognition model that OpenAI released last year.

When it launched in December, ChatGPT was powered by a neural network called GPT-3.5. OpenAI has since upgraded the chatbot service to its latest GPT-4 model. Compared with its predecessor, the latter model can interpret user instructions more accurately and is less prone to AI hallucinations.

As part of its internal software testing efforts, OpenAI had GPT-4 complete a simulated version of the U.S. bar exam. The chatbot earned a score that put it in the top 10% of test takers. It also completed dozens of other academic tests spanning fields such as mathematics, law and economics.

Another major enhancement in GPT-4 is that it can take images as input. A user could, for example, upload a photo of a street sign and ask the model to translate the text it contains. GPT-3.5, the model that powered the initial version of ChatGPT, didn’t offer such a capability. 

Although the Android and iOS versions of ChatGPT are both based on OpenAI’s latest GPT-4 model, they don’t yet support image input. The startup may roll out the capability with a future update to its mobile apps.

It’s also possible the mobile apps will eventually receive support for ChatGPT Business, a paid edition of the chatbot that OpenAI previewed earlier this year. It’s designed for enterprises that require user management features and professionals who “need more control over their data.” In April, OpenAI stated that ChatGPT Business was set to launch within a matter of months. 

Image: Unsplash