Following a two-month period of exclusivity for Apple users, Android users can now enjoy the OpenAI's ChatGPT app on their mobile devices.

On Tuesday, July 25 Eastern Time, OpenAI announced that users in the United States, India, Bangladesh, and Brazil can now download the Android version of the ChatGPT app from Google Play and provided the relevant links. OpenAI mentioned that it plans to expand the app's availability to more countries next week.

When OpenAI launched the iOS version of the ChatGPT app in May this year, they promised that an Android version would soon follow. As indicated by OpenAI last week, the Android version of ChatGPT arrived as scheduled.

The ChatGPT app allows users to sync their history across multiple devices and supports voice input. Users can use the app to ask ChatGPT questions for real-time answers, get suggestions from ChatGPT, or even have ChatGPT help compose emails or texts for PowerPoint presentations.

The arrival of the Android version coincides with a decline in the explosive growth of ChatGPT users, with both download and visit volumes on a downward trajectory.

Previously, SensorTower data cited by Wall Street Journal showed that in June this year, the installation volume for ChatGPT and Microsoft's Bing app dropped by 38%. Meanwhile, SimilarWeb data revealed that traffic to the ChatGPT website across desktop and mobile platforms declined 9.7% month-over-month in June. The number of unique visitors to the website also fell by 5.7%, and the average time spent by each user on the site decreased by 8.5%.

Moreover, the latest iteration of GPT has been criticized for seemingly "getting dumber." Some users noticed last month that the flagship model, GPT-4, appeared less intelligent. Recent studies have suggested that GPT-4's mathematical ability and coding competence have both deteriorated.

Some experts in the field of language models speculate that OpenAI is developing several smaller GPT-4 models, each behaving similarly to the larger model but with lower operational costs, to improve cost-effectiveness using a blended expert model (MOE) approach.

OpenAI responded by stating that since the launch of GPT-4 on March 14, the model itself has remained static, implying that the model was not contaminated by extensive external data. However, they admitted the model has inherent instability, resulting in inconsistent answers to similar prompts. OpenAI added that it would continue to update its API.

Meanwhile, as the competition in AI continues to heat up, OpenAI faces emerging rivals. Last week, one of OpenAI's main funders, Microsoft, partnered with Meta to launch a commercial version of the open-source AI model Llama, offering businesses an alternative to models provided by OpenAI and Google.