Batch resize
Many images at once · by width / height / fixed size / percentage · export PNG / JPG / WebP in one download
缩放参数
输出文件名：
原名-宽x高.扩展名
图片列表
0 张添加图片后在这里显示：原尺寸 → 新尺寸，生成后可逐张下载。
FAQ
No. Resizing is resampling by interpolation: shrinking throws pixels away, enlarging guesses new ones - neither adds detail the picture never had. The tool is meant for shrinking and unifying sizes (an 180x180 avatar, a 200x200 thumbnail); for a real enlargement that invents detail, reach for a dedicated AI upscaling tool.
Stretch fills the given width and height outright and squashes the picture when the ratios differ. Pad (scale proportionally) keeps the whole frame and its proportions, filling the leftover area with a background color - transparent, white or black; PNG and WebP support transparency. Crop (scale proportionally) grows the picture until it fully covers the frame, then trims the excess from the center - no distortion, but the edges get cut away.
The JPG format has no transparency channel - pick a transparent background and the JPG export silently switches to a white underlay; for real transparency choose PNG or WebP. Writing GIF needs per-frame animation encoding, which this tool does not produce: a GIF input is processed as its first frame and exported as PNG, transparency kept.
Nothing. Decoding, resizing and encoding all run in your browser through canvas; the pictures never leave the machine, and reloading clears everything. Up to 30 pictures per batch, at most 20MB each, source long edge at most 16384 pixels; output edges are kept within 1-16384 pixels too.