JP Morgan praises Google: AI technology is leading, far surpassing previous innovation speeds, "AI reshaping search" has taken shape, and monetization has also been considered

Google is undergoing a "complete revolution in search," which may be the biggest change since the birth of search engines.

After the Google I/O 2025 Developer Conference, JP Morgan highly praised Google's leading position in artificial intelligence (AI), especially regarding the Gemini model.

On the 22nd, according to news from the Chasing Wind Trading Desk, JP Morgan analyst Doug Anmuth's team believes that Google is releasing products at an unprecedented speed. Google's full-stack AI strategy encompasses world-class research, custom AI hardware, robust global infrastructure, leading foundational models, and developer tools and platforms.

JP Morgan particularly emphasized the "AI Mode," considering it a significant advancement for Google in the search field. Google's foundational models are also leading in efficiency and cost, and Google has begun to consider commercializing AI technology.

JP Morgan maintains an "Overweight" rating on Google stock, with a target price of $195. As of Wednesday's close, Google's stock price rose 2.87% to $170, leaving a 14.71% upside to the target price.

Google's Innovation Accelerates: AI Product Implementation Speeds Up

JP Morgan believes that the information conveyed from the Google I/O conference indicates that Google is releasing products at an unprecedented speed.

Especially in terms of foundational models, the Gemini model is in a leading position. AI Mode has integrated Gemini into search functionality and incorporated the capabilities of Astra, Mariner, and Deep Research. The launch of AI Mode in search is only one year later than the AI Overview.

JP Morgan stated that Google has adopted a full-stack AI approach, with world-class research, custom AI hardware, robust global infrastructure, leading foundational models, and developer tools/platforms. This gives Google's foundational models an advantage in efficiency and cost.

Data shows that currently, Google's products and APIs process 480 trillion tokens per month, which is 50 times the 9.7 trillion tokens/month in 2024. Meanwhile, the number of developers using the Gemini API has exceeded 7 million, growing fivefold since the last I/O conference.

"Comprehensive Restructuring of Search": AI Mode Completely Transforms Search Experience

JP Morgan particularly emphasized "AI Mode," describing it as "the comprehensive restructuring of search" and "the biggest revolution since the invention of search engines."

AI Mode provides an end-to-end AI search experience, leveraging connections with the Google product suite, advanced reasoning, and multimodal capabilities to achieve more complex queries (2-3 times longer than traditional searches), deeper personalized insights, and follow-up questions.

Additionally, AI Overviews have expanded to over 1.5 billion monthly active users, across more than 200 countries and regions, and in over 40 languages, becoming one of Google's most successful search product launches in the past decade. In Google's largest markets, such as the United States and India, the usage of query types showcasing AI overviews has increased by over 10%.

JP Morgan noted that upcoming enhancements include:

1. In-depth research capabilities for expert-level reports
2. Real-time visual interaction features of Project Astra
3. Agent capabilities of Project Mariner, allowing actions and automation tasks in the browser
4. Enhanced shopping features, including virtual try-ons and improved product discovery
5. Greater personalization and integration with other Google applications
6. Complex and customized data analysis and visualization

Commercialization and Ecosystem: Gemini Applications and Subscription Model

Google also announced new improved versions of Gemini, Gemma, and generative media models. Within the Gemini family, Google announced upgrades for Gemini 2.5 Pro and Gemini 2.5 Flash, as well as an experimental demonstration of its new text model, Gemini Diffusion.

Notably, Gemini 2.5 Pro and Flash ranked first and second in the LM Arena, with 2.5 Pro also leading in other rankings such as LaMDA and WebDev Arena.

For developers, Google announced Gemma 3n, an open-source multimodal model capable of running on devices. In terms of generative media models, Google announced the newly improved image and video generation models Imagen 4 and Veo 3.

JP Morgan stated that Google showcased its commercialization intentions at the I/O conference. At this conference, Google announced several enhancements for the Gemini application, including:

1. Gemini Live, offering camera and screen sharing capabilities, with integration with other Google applications in the coming weeks
2. Google's latest image and video generation models Imagen 4 and Veo 3
3. In-depth research with personal PDF and image upload capabilities
4. Canvas equipped with improved coding from Gemini 2.5 Pro
5. Gemini-Chrome integration, which can clarify or summarize information on any webpage At the same time, Google launched a new premium subscription tier, Google AI Ultra ($249.99/month), which provides access to Google's most powerful models and advanced features. The $19.99/month Advanced tier has been upgraded to Google AI Pro, offering better functionality at the same price