Agent Studio provides a centralized, collaborative workspace for discovering models, refining system instructions, and optimizing prompts. By integrating built-in experiences with the broader Google Cloud ecosystem, Agent Studio helps you transform innovative ideas into production-ready generative AI solutions.
Key capabilities
- Interactive canvas view: A dual-pane workspace that lets you preview and edit code for the artifacts you are developing, ensuring a seamless transition from visual design to implementation.
- Prompt refinement and comparison: Use built-in natural language tools to automatically generate system instructions and perform side-by-side comparisons to test how different instructions and models affect agent behavior.
- Efficient developer workflow: Navigate large prompts with a summarized history minimap. Use slash commands or ask for help using Google Cloud assistance to manage tools, handle conversation context, and generate sample code quickly.
- Grounding and RAG: Enhance response accuracy by grounding agents in real-time data from Google Search and Google Maps, or integrate your own enterprise data using RAG Engine, Agent Search, or Elasticsearch.
- Granular control and settings: Fine-tune model performance with adjustable thinking levels, structured output requirements, and customizable safety filters.
Agent Studio offers advanced model access. Build with Google's most capable models, including Gemini 3.1 for reasoning, Nano Banana 2 for low-latency image generation, and Veo 3.1 for high-quality video generation.
Experience seamless integration with Agent Studio's cloud-native integration with the broader Google Cloud ecosystem, bringing you access to APIs and deployments.
What's next
Agent Studio quickstart: Send text prompts to Gemini
Learn how to send text prompts to Gemini using Agent Studio.
Deploy your prompt as a web application
Follow the guide to deploy your prompt as a web application from Agent Studio.