Home / Technology / New Google AI Agent Embeds Research in Apps
New Google AI Agent Embeds Research in Apps
12 Dec, 2025
Summary
- Google launched Gemini Deep Research, powered by Gemini 3 Pro.
- The new agent integrates Google's research capabilities into apps via Interactions API.
- Google faces strong competition with OpenAI's newly released GPT 5.2.
Google has unveiled a significantly updated version of its Gemini Deep Research agent, leveraging the advanced Gemini 3 Pro foundation model. This new iteration moves beyond generating research reports, enabling developers to embed Google's sophisticated research functionalities directly into their own applications via the novel Interactions API.
The Gemini Deep Research agent is engineered to process and synthesize extensive information dumps. Its capabilities are already being utilized by customers for critical tasks, including due diligence and assessing drug toxicity. Google plans to integrate this powerful agent across its services, such as Google Search and Finance, signaling a shift towards AI-driven information retrieval.
This development occurs amidst a heated AI landscape, with Google launching its update on the same day OpenAI announced its GPT 5.2 model. Benchmarks show Google's agent performing strongly, though OpenAI's latest offering is a close competitor, even surpassing Google in certain browser-based tasks.
