https://opal.google
- App Generation: An LLM interprets app ideas described in plain English and automatically generates the visual workflow of nodes and connections.
- The "Generate" Step: The core building block of an Opal app is the Generate node. Specific Google AI models (like Gemini) can be selected inside this node to process text, images, or data.
- Prompt Chaining: Prompts can be "chained" so the output of one LLM call becomes the input for the next.
- Agentic Intelligence: Recent updates have introduced "Agent" steps. These agents use LLM reasoning to proactively decide which tools (like Google Search) or models (like Veo for video) are needed to complete a task.
