Google Opal No-Code Agent Builder Ships Goal-Driven Workflows: AI Agents That Select Tools and Initiate Conversations Autonomously

VentureBeat·medium signalXBlueskyLinkedInCopy link

Google's Opal, an experimental no-code tool from Google Labs announced at I/O 2026, chains together AI model calls, tools, and prompts into visual workflows without generating code. Builders can now define a goal and let the agent autonomously select tools (including Gemini 3 Flash, Gemini 3.1 Pro, and Veo for video generation) and initiate conversations with users when it needs more information. VentureBeat called it 'the new blueprint for building AI agents' in enterprise settings, noting it could democratize agent creation for non-engineers in a way that framework-level tools like LangGraph cannot.