Launching Autonomous Image Generating Agents

This guide explains how to create, publish, and use Image generating agents on the Syntax platform. It also covers best‑practice prompting and how to choose the right image generating Models and Styles for launching agents that produce consistent, high‑quality results.

---

What is an Image Generating Agent?

An Image Generating Agent is your own AI image generator with a fixed style and behavior, which can be used by you or anyone on the Syntax platform to geenrate images. You define it by choosing:

• a Base Model (an SDXL image model, which is the core model responsible for generating all images),

• up to 3 Image Styles (these are Low Rank Adaptations based on your model choice, further specializing the style of images your agent can create), and

• a Base Prompt (this is the core instruction that always serves as context for the agent to generate any image).

Agents are deployed on‑chain. It costs 0.001 ETH to create an agent and launch your agent on Base. After creation, the agent has its own ERC‑20 token. Those tokens are used by anyone to pay for image generations with that agent. As the creator, you can buy up to 5% of your agent’s tokens (inside the app) to generate images yourselves or to seed liquidity.

---

Browse Existing Agents or Create your Own

Head to explore agents on the Syntax platform to see a gallery of image generatin agents, and pick up one you like to generate your images. Or you can choose to build your own!

Dive deep into more information on how to launch and create agents:

1. Launch your own Agent

   1. Key Concepts

   2. Quickstart

   3. Creating an Agent (Step by Step)

      1. Designing a strong base prompt

      2. Choosing the right base model

      3. Prompting for the best results (for you and your users)

      4. Troubleshooting

      5. Agent Quality Checklist

2. Browse Existing Agents

3. Tokenomics