Image AI Models Are Now the Biggest Growth Engine for Apps and the Numbers Are Staggering
Something has quietly shifted in the AI app race, and the data makes it impossible to ignore. Image AI models are now the single biggest driver of app growth, outpacing every chatbot upgrade the industry has thrown at users. That is the finding from a new report by app intelligence firm Appfigures, which tracked download trends across major AI platforms and found that visual model launches generate 6.5 times more downloads than standard model updates.
Not long ago, it was text-based breakthroughs that moved the needle. A new conversational model or a slicker voice interface was enough to pull millions of new users in. That era appears to be over. Users today are responding to what they can see, and that hunger for visual AI capabilities is reshaping how the biggest players in the industry compete for attention.
The clearest example comes from Google. When Gemini released its image model Nano Banana following the Gemini 2.5 Flash image model launch, the app racked up more than 22 million additional downloads in the 28 days that followed, a jump of over 4x compared to its normal baseline. ChatGPT was not far behind. The release of OpenAI’s GPT-4o image model in March 2025 brought in more than 12 million incremental installs over the same 28-day window. That figure was approximately 4.5 times higher than what ChatGPT saw when it launched GPT-4o, GPT-4.5, and GPT-5 as text-focused model upgrades.
The pattern held for Meta too, though at a smaller scale. The launch of Vibes, Meta AI’s AI video feed, added an estimated 2.6 million incremental downloads in the 28 days after its September 2025 debut. While technically a video feature, it sits firmly in the world of visual AI, and the download response tracked accordingly.
What the data makes clear is that image AI models drive app growth in a way that text upgrades simply no longer do. But there is a catch that the industry has not yet solved.
Downloads and dollars are not the same thing. Gemini’s Nano Banana launch generated a massive spike in installs but produced only an estimated $181,000 in gross consumer spending during that same 28-day period. Meta’s Vibes drew new users in but brought in no meaningful revenue at all. Only OpenAI managed to turn the moment into money. ChatGPT’s 4o image model release generated an estimated $70 million in gross consumer spending over 28 days compared to its prior baseline, a result that separates it from the rest of the field.
The Appfigures report also examined DeepSeek, though it stood apart from the others. The DeepSeek R1 release in January 2025 drove 28 million downloads, but that surge came from sheer novelty and the shock of a low-cost AI model shaking the entire industry, not from image capabilities. That kind of curiosity-fueled growth is a different beast entirely.
The takeaway is that image AI has become the most powerful acquisition tool in the AI app market right now. The harder challenge, as most of these platforms are learning, is converting that wave of curious downloaders into paying customers.
