That sounds like a great approach! For generating visuals you can customize, tools like MidJourney, DALL·E, and Stable Diffusion are really popular. They let you guide the style and details, so you can tailor outputs to your project’s theme. Stable Diffusion, in particular, is very flexible if you want to run it locally and tweak prompts heavily.
If you’re also thinking about text-to-image ideas or brainstorming concepts, ChatGPT can be surprisingly useful—it can help you craft detailed prompts for these image generators. In China, since direct access to ChatGPT might be blocked by the GFW, people often use
稳定 ChatGPT 镜像站, VPNs, or curated GitHub lists of working mirrors. This way, you can still get the AI assistance you need without running into accessibility issues.
Combining ChatGPT for prompt creation with a visual AI for execution can really speed up the design process while keeping your ideas unique.