Fetch.ai has announced a guide for creating a Google Gemini Image Generation Agent, which allows users to generate images from text prompts using Google’s Gemini 2.5 Flash Image model. This integration leverages the chat protocol for seamless communication between agents and utilizes Agentverse’s ExternalStorage for sharing generated images, ensuring efficient management of binary assets without bloating message data. The guide provides detailed steps, including setting up the project environment and registering the agent on Agentverse, enabling developers to embed this image generation capability into their mailbox agents.
Fetch: Fetch is the core project and framework from Fetch.ai used to create autonomous agents with mailbox and chat protocol support. The tutorial relies on Fetch’s uagents library and ExternalStorage service to handle image uploads, permissions, and resource sharing. Fetch enables the agent to run locally while remaining discoverable and interoperable within the broader Agentverse marketplace.
Gemini: Gemini refers to Google’s family of multimodal AI models, with the 2.5 Flash Image variant specifically designed for generating images from text prompts. The news demonstrates direct integration of this model into a Fetch.ai agent via the google-genai library to produce and share visual outputs. Gemini’s image generation is presented as the key capability that the custom agent exposes to other agents and users on Agentverse.
Google: Google is a multinational technology company that develops AI models and tools including the Gemini family of large language models. In this news, Google supplies the Gemini 2.5 Flash Image model that powers the image generation capabilities of the custom agent. The tutorial shows how developers can connect Google AI APIs directly to Fetch.ai mailbox agents running on Agentverse.
Fetch.ai: Fetch.ai develops the uagents framework and Agentverse platform for building, deploying, and discovering autonomous AI agents. The company’s tools form the core of the provided tutorial, enabling the Gemini image agent to handle chat protocols, external storage, and registration for use with ASI:One. Fetch.ai’s ecosystem supports multi-agent collaboration and mailbox functionality highlighted throughout the guide.
Agentverse: Agentverse is Fetch.ai’s global marketplace and hosting platform where agents can be registered, discovered, and connected with users or other agents. The news instructs developers to register the Gemini image agent on Agentverse so it becomes accessible via ASI:One and the Agent inspector. Agentverse provides the external storage backend required for securely sharing generated images.
ASI:One LLM: ASI:One LLM is the conversational interface that allows users to discover, select, and interact with registered Agentverse agents through natural language. The tutorial explains how to enable agent mode in ASI:One so that image generation requests are routed to the custom Gemini agent. ASI:One serves as the primary testing and usage front-end for the completed agent.
`json
{
“Agent Integration”: “Google Gemini models can be integrated into Fetch.ai mailbox agents to enable image generation capabilities that are shared through the chat protocol.”,
“Platform Features”: “Agentverse ExternalStorage enables agents to upload, set permissions, and reference binary assets like generated images without embedding them directly in messages.”,
“Developer Workflow”: “The step-by-step guide demonstrates how to integrate Google AI APIs with Fetch.ai’s uagents and chat protocol for creating tools that are discoverable and compatible with multiple agents.”
}
`