Google has upgraded its Gemini Deep Research agent for consumer-facing apps and developers. The new Interactions API allows for advanced tool use and extended reasoning, going beyond basic text generation. Developers can now connect Gemini models, internal agents, and custom-created agents through a single API. The Deep Research agent, powered by Gemini 3 Pro, excels in long-running tasks and web navigation, providing highly specific information and enhanced search accuracy. Tests show significant improvement over the standard Gemini 3 Pro model, making the agent more capable of handling complex research tasks. Developers can preview the upgraded capabilities on Google AI Studio, with a rollout planned for consumer products like the Gemini app, Google Search, and NotebookLM.

Read more at Nasdaq: Google Rolls Out Upgraded Gemini Deep Research And Opens Agent Access To Developers