Can I use the guide page directly?
Yes. The workbench is built into this page, so you can upload an image and generate a prompt while reading the guide.
Image to Prompt Guide
Learn the workflow, then try it on your own image
Use this page as a practical guide: upload a reference, generate a prompt, inspect the structure, then continue into image or video generation without leaving the workbench.
1
Upload a reference or paste an image URL
Start with the visual you want to understand. The tool reads subject, scene, lighting, style, and composition from the image.
2
Choose the output you actually need
Use plain prompt, structured view, or JSON depending on whether you want quick copying, editing control, or machine-readable output.
3
Send the prompt into the next tool
After the prompt is ready, continue into text-to-image, image-to-image, text-to-video, or image-to-video from the same workspace.
Try it here
Start with your own reference image
The same workbench from the homepage is available here, so you can follow the guide and test the tool in one place.
点击、拖放或粘贴图片到这里
支持 PNG、JPG 和 WebP，最高 10 MB。
也可以直接从剪贴板粘贴图片
输出风格
从图片开始
生成提示词、结构化视图或 JSON。
Guide FAQ
What to know before you generate
Is this only for Midjourney?
No. You can generate general prompts, Midjourney-style prompts, Flux-friendly prompts, Stable Diffusion prompts, and JSON output.
What should I do after the first prompt?
Copy it, edit a specific part, or send it directly into image or video generation using the tools in the workbench.