Google has restricted free access to its Nano Banana Pro image generation model, limiting non-paying users to two images per day, down from three, due to high demand as outlined in a support document.
The company updated its policies in response to surging usage of the AI system, which enables users to create and edit images. A support document, first identified by 9to5Google, specifies that free users can generate two images daily under the new restrictions. This change applies specifically to the Nano Banana Pro model, introduced as a tool for producing various visual content during periods like the current long holiday weekend. Google attributes the adjustment to elevated interest in these features.
Directly from the document, Google states, “Image generation and editing is in high demand.” The policy further clarifies that these limits may change frequently and will reset daily, allowing users a fresh allocation each day. This mechanism ensures ongoing access while managing server load and resource allocation for the AI infrastructure supporting the model.
In parallel, Google has imposed limitations on free usage of Gemini 3 Pro, another advanced AI offering. Non-paying users receive basic access, with daily limits subject to frequent modifications. The rollout of Gemini 3 Pro began on November 18, when the company committed to providing five free prompts per day for non-subscribers. This initial limit aligned exactly with the provision for Gemini 2.5 Pro, maintaining consistency across models at launch.
Subscribers to paid tiers face no alterations to their allowances. The Google AI Pro plan continues to permit 100 prompts per day, while the AI Ultra plan allows up to 500 prompts daily. These higher thresholds cater to users requiring extensive interaction with the AI systems for professional or intensive applications.
Google’s actions follow a pattern observed in the AI sector. OpenAI encountered similar challenges when introducing its built-in image generator for ChatGPT. The company postponed availability to free users after demand exceeded projections, straining computational resources. Despite the delay, OpenAI proceeded to extend image generation capabilities to non-paying accounts once capacity stabilized.
