Our Take
Downloads spike hard for image features, but only OpenAI converts curiosity into cash.
Why it matters
Product teams betting on visual AI for growth need conversion strategies, not just capability launches. The pattern shows users will try image generation but won't pay without clear ongoing value.
Do this week
Product managers: Audit your image AI conversion funnels this week so you can capture revenue during the next visual model launch window.
Visual AI drives 6.5x more installs than text models
Image model releases generate 6.5x more app downloads than traditional conversational AI updates, according to Appfigures analysis of major AI mobile apps. Google's Gemini added 22 million downloads in the 28 days after launching its Nano Banana image model in August, lifting downloads by more than 4x (per Appfigures). ChatGPT pulled 12 million incremental installs after introducing GPT-4o image capabilities in March, roughly 4.5x more than its text-only GPT-4o, GPT-4.5, and GPT-5 releases combined (company-reported data).
Meta AI's video feed feature Vibes added 2.6 million downloads in 28 days after its September 2025 launch. DeepSeek R1 drove 28 million downloads in January 2025, though this reflected the company's breakout moment rather than image-specific interest.
Download spikes rarely convert to revenue
The download surge doesn't translate to subscription growth for most apps. Gemini's Nano Banana generated only $181,000 in estimated gross consumer spending during its 28-day launch window despite producing the largest download spike. Meta AI's Vibes launch drove downloads but no meaningful revenue.
Only ChatGPT converted attention into dollars, generating an estimated $70 million in gross consumer spending over 28 days after its 4o image model launch (per Appfigures estimates). This suggests the revenue challenge isn't inherent to image AI but reflects execution differences in conversion strategy and product positioning.
Focus conversion, not just capability
The pattern reveals a consistent user behavior: people install apps to test visual AI capabilities but don't stick around as paying customers. Product teams should prepare conversion funnels before launching image features, not after the download spike peaks.
OpenAI's success suggests subscription conversion requires more than impressive demos. The revenue gap between ChatGPT and competitors points to onboarding, retention mechanics, and ongoing value delivery as the real differentiators in visual AI adoption.