Google is limiting access to its most powerful AI models for free users and even some paying subscribers. The company announced today that free users of its Gemini AI chatbot will soon only be able to use the Gemini 1.0 Pro model, which is the weakest of the three models currently available.
Currently, free users can choose between Gemini 1.0 Pro, Gemini 1.5 Pro, and Gemini Advanced (powered by Gemini Ultra 1.0). Starting July 13, only Gemini 1.0 Pro will be available without a subscription.
The change also impacts Google One AI Premium subscribers. They will lose access to Gemini 1.5 Pro and will be limited to Gemini 1.0 Pro unless they upgrade to the AI Premium plan, which costs $20 per month.
Gemini 1.5 Pro is significantly more capable than Gemini 1.0 Pro, particularly when it comes to processing large amounts of information. Google introduced Gemini 1.5 Pro in February, touting its ability to analyze 1 million tokens of text or code. Gemini 1.0 Pro, by comparison, can only handle 32,000 tokens.
“We’re making these changes to provide a better experience for all users and to ensure the long-term sustainability of Gemini,” a Google spokesperson said in a statement. “We believe that limiting access to our most powerful models will help us to maintain the quality of our service and to continue investing in AI research and development.”
The move comes as Google faces increasing competition from other AI companies, such as OpenAI and Anthropic. OpenAI offers access to its GPT-4 model through a paid subscription, while Anthropic offers access to its Claude 3 model through a similar model.
Google’s decision to limit access to its most powerful AI models is likely to be controversial. Many users have come to rely on Gemini 1.5 Pro for tasks such as summarizing long documents, writing code, and generating creative content. Limiting access to this model could make it more difficult for users to accomplish these tasks.