<a href="https://www.google.com/preferences/source?q=finance.biggo.com” rel=”nofollow noopener” target=”_blank”>Add preferred source
Google gathered AI app developers from around the world in Singapore for its “AI App Day” event, unveiling the explosive growth of the AI-native app market and its support strategy. AI app payments doubled in a single year, while the B2B agent segment posted an annual growth rate of 633%. While simple chatbots are seeing slowing growth, the interactive entertainment genre is recording daily usage times of up to 146 minutes. To address the high inference costs of generative AI apps, Google proposed a hybrid monetization model combining subscriptions, in-app purchases, and advertising, along with a full-stack support strategy linking AI models, cloud infrastructure, app distribution, and advertising. Google Play also previewed short-form video-based app discovery features and a prepaid pay-as-you-go billing system.
Key Elements
Google brought together AI app developers from around the world in Singapore to showcase the growth of the AI app market, which is expanding beyond chatbots into interactive entertainment and business-to-business (B2B) segments. AI app market payments doubled in a single year, while the B2B agent segment recorded an annual growth rate of 633%.
The “Google AI App Day” event, held on the 17th (local time) at the Andaz Hotel in Singapore, drew 195 founders from roughly 140 app development companies worldwide. The event was organized to drive growth in AI-native apps and short-form drama content while exploring strategies for global business success. AI-native apps refer to next-generation applications built with artificial intelligence as the core pillar from the planning and design stages, rather than simply bolting AI features onto existing software.
Chatbots Slow, Interactive Entertainment Reaches Up to 146 Minutes Daily
Paula Wang, managing director of Google Play partnerships for Asia-Pacific, assessed that the global AI app ecosystem is evolving from simple Q&A tools into emotionally engaging, immersive experiences.
“After tracking the AI-native app market over the past two years, we’re seeing growth slow for AI chatbots and assistants that simply answer questions,” she said. “The interactive entertainment genre, where users create their own stories, is approaching average daily usage times of 60 to as much as 146 minutes.”
Growth in the B2B AI app segment is also notable. “The B2B agent segment is posting an annual growth rate of 633%,” Wang said. “The base of AI-native apps is still small, but it’s growing very rapidly without significant industry support.” She added, “Just as web-centric services previously migrated to mobile apps, the boundaries between apps, games, and content creation are now dissolving.”
Mecca Okereke, vice president and general manager of Google Play apps, emphasized, “We are in a period of transformation unprecedented in computing history. Google Play will work with developers across the entire lifecycle—from building innovative AI services to acquiring users and generating long-term revenue.”
Lower Barriers to Entry, But Inference Costs Squeeze Profitability
While AI has made app development easier, surviving in the market has arguably become harder. Karen Teo, vice president of platform and device partnerships for Google Asia-Pacific, told reporters, “Anyone can now launch an AI-powered app. Users won’t just use any app—they’ll only spend time on services with quality and creativity.”
The high cost structure of AI-native apps is another hurdle developers must overcome. Traditional software doesn’t see significant additional costs as users grow, but generative AI-based apps require computing power to process prompts. According to Google, inference costs for these apps can account for up to 35% of revenue. The more free users engage with AI features, the greater the cost burden, deteriorating profitability.
As a result, companies operating generative AI-based apps have average gross margins of 50–60%, lower than the 80–90% typical of traditional software-as-a-service (SaaS). During periods of rapid user growth, margins can fall to 25–30%.
Aman Grover, director of app development sales for Google Asia-Pacific, stressed, “Unlike traditional software, generative AI apps see GPU inference costs account for 20–35% of total revenue, meaning losses grow as usage increases. Developers need to move away from chasing download numbers and shift toward revenue-focused marketing based on advertising return on investment.”
Google’s Solution: Hybrid Monetization and Full-Stack Support
On the revenue model front, Google proposed “hybrid monetization” that combines subscriptions, in-app purchases, and advertising. The approach involves showing ads to free users to recover some inference costs, while selling additional usage in the form of credits.
On the infrastructure side, Google unveiled a “full-stack” strategy that bundles its services together. By connecting Google DeepMind’s AI models, Google Cloud’s infrastructure, Google Play’s distribution network, and Google Ads and AdMob’s user acquisition and monetization capabilities, Google aims to reduce the burden on developers who would otherwise need to separately handle model development, infrastructure buildout, global distribution, user acquisition, and monetization.
Manish Gupta, senior director of research at Google DeepMind, said, “For the Asia-Pacific market, where diverse languages and cultures coexist, we’ve introduced the Gemini 3.5 translation model with speech recognition supporting more than 85 languages.” He added, “The Gemma family of lightweight on-device models improves inference complexity, delivering up to 2.5 times better token efficiency compared to competing models.”
Bill Jia, vice president of engineering at Google Cloud, said, “Google Cloud’s seventh-generation Tensor Processing Unit (TPU), our AI accelerator, delivers 50% better performance per dollar compared to previous generations. Capcom, for example, has adopted Gemini Enterprise agents to save more than 30,000 hours of game testing engineering resources each month.”
Google Play’s app discovery and payment systems are also being revamped. Austin Shoemaker, director of product management at Google Play, said, “We’ve introduced Play Shorts, a swipeable short-form video feature at the top of Google Play for browsing apps, and a feature in the Android Gemini app that lets users find and install apps through conversation.” Considering the high inference costs of AI apps, he added, “We’ll soon launch a prepaid pay-as-you-go system that automatically tops up credits when usage limits are exhausted.”
“Why Users Spend More Time on AI Apps Than Netflix”
Kim Jong-yoon, CEO of South Korean social AI startup Scatter Lab, participated in a panel session at the event. Scatter Lab operates the AI chat app “Zeta” in the South Korean, U.S., and Japanese markets.
“The reason users spend more time on AI apps than on Netflix is that they’re consuming real-time interactive storytelling, beyond simple conversation or emotional exchange,” Kim said. “We’ve lowered inference costs through in-house lightweight model optimization and introduced in-app advertising, including full-screen ads for free users.”
He cited Google’s global market experience and full-stack support as key strengths, saying, “Google has powerful experience in global markets, and there’s a lot for South Korean companies to learn as they expand worldwide. We’ll continue this partnership.”
Once added, BigGo Finance appears first in Google Search Top Stories, so you get the broadest, most up-to-the-minute, and most comprehensive global financial news first.
