Image model releases spark massive growth for AI mobile apps

A comprehensive report from app intelligence provider Appfigures has revealed a significant shift in the mobile landscape, as image model releases now drive growth six times more effectively than traditional updates.

This data suggests that the era of conversational text dominance is being eclipsed by a surge in demand for visual capabilities.

Google’s Gemini experienced a remarkable transformation following the introduction of the Gemini 2.5 Flash image model, Nano Banana, last August.

The launch propelled the application to more than 22 million additional downloads within twenty-eight days, representing a fourfold increase.

Similarly, OpenAI’s ChatGPT secured over 12 million incremental installs following the debut of its GPT-4o image model in March last year.

This figure vastly outperformed the demand generated by previous iterations, including GPT-4.5 and GPT-5.

However, the report noted that "additional downloads don’t always translate into increased mobile revenue." While Gemini saw a massive spike in user acquisition, it generated only $181,000 in gross consumer spending during that window.

In contrast, OpenAI capitalised on the momentum, with GPT-4o leading to an estimated $70 million in revenue.

Meta AI also followed this trend on a smaller scale, with its "Vibes" video feed adding 2.6 million downloads in September 2025.

DeepSeek remains a notable outlier in the analysis; its R1 model drove 28 million downloads in January 2025.

This breakout success was attributed to industry curiosity regarding cost-effective training techniques rather than specific image features.

Looking ahead, developers are expected to focus heavily on these visual tools to maintain high user engagement.