Designs that lost their layers get their structure back
The client sent only the final JPG and nobody has the source file. img2htm analyzes the visible result and rebuilds the relationships between headings, body copy, images and tables. It does not recover the original source; it creates a visually faithful, semantically structured web representation.
Works with your design system
Output CSS is organized with i2h- prefixed classes so you can map it to your own tokens. Separated image assets are delivered as individual PNGs.
Figma and Photoshop workflows
See the Figma and Photoshop entries in the integrations directory to convert straight from your canvas.
核心优势
- Semantic HTML drafts generated automatically
- Separated image and icon assets
- CSS organized with prefixed classes
- Responsive reflow by default
- Convert straight from Figma or Photoshop
- Works without the original file
核心优势
Semantic HTML drafts generated automatically
Separated image and icon assets
CSS organized with prefixed classes
Responsive reflow by default
Convert straight from Figma or Photoshop
Works without the original file
工作原理
上传图片
上传 JPG、PNG 或 WebP，最大 25 MB。每 15,000 纵向像素消耗 1 点。
AI 分离与重构
理解标题、正文、价格、表格和列表的层级，并还原为真实文本。
HTML 校验并发布
生成在桌面、平板和手机上自然重排的语义化 HTML 与 CSS。