← Back to Tools
📐

Image Resizer

Resize to any dimension or social preset

Loading tool...

Social media size cheat sheet (2026 edition)

Every platform automatically compresses and crops uploads that don't match their expected dimensions. Use the presets built into the resizer to hit the exact target first time:
PlatformTypeDimensions
InstagramSquare Post1080 × 1080 px
InstagramPortrait Post1080 × 1350 px
InstagramStories / Reels1080 × 1920 px
YouTubeThumbnail1280 × 720 px
LinkedInBanner1584 × 396 px
X (Twitter)Header1500 × 500 px
X (Twitter)Timeline Image1200 × 675 px

How the resizer works under the hood

The tool draws your image onto an HTML Canvas element at the target dimensions using Lanczos-3 resampling. This is a sinc-based interpolation filter with a 3-lobe window that calculates each new pixel by sampling a weighted neighbourhood of surrounding source pixels.

Why does this matter? Compare the three common algorithms:

    1. Nearest-neighbour: Fast, but produces blocky staircasing artifacts on diagonal edges.
    2. Bilinear: Smooth, but blurs fine details — especially bad for screenshots with text.
    3. Lanczos-3: Preserves sharpness on downscale by applying a mathematically optimal weighting curve. Text and line art stay crisp; photo edges stay clean.
The Canvas API handles Lanczos-3 in milliseconds for typical photo sizes, so you get studio-quality resampling with zero waiting.

How to use the Image Resizer

  1. Upload your image — Drop a PNG, JPEG, or WebP file. The tool reads and displays the original dimensions before you make any changes.
  2. Set target dimensions — Type in pixel values, or select a social media preset. Toggle 'Lock Aspect Ratio' to scale proportionally.
  3. Resize and download — The canvas renders the resampled image. Click Download to save in the same format as your source.

Tips for best results

    1. Downscaling: Always better quality than upscaling. If you have a 4K photo and need a 1080×1080 thumbnail, the Lanczos output will be sharp.
    2. Upscaling small images: No algorithm can invent pixels that don't exist. For 2× enlargements or more, consider using a dedicated AI upscaler first.
    3. PNG logos: Aspect ratio lock is critical — a 512×512 icon resized to 200×150 without lock will look squished. Always keep the lock enabled for logos.

FAQs about Image Resizer

Everything you might be wondering — answered.

What resampling algorithm is used?+
Lanczos-3 (a sinc-based interpolation filter with a 3-lobe window). It produces sharper results than bilinear interpolation when downscaling, at the cost of slightly more computation — which the Canvas API handles in milliseconds for typical photo sizes.
Which social media presets are included?+
Instagram (Square 1080×1080, Portrait 1080×1350, Landscape 1080×566, Stories 1080×1920), YouTube (Thumbnail 1280×720, Channel Art 2560×1440), LinkedIn (Profile 400×400, Cover 1584×396, Post 1200×627), X/Twitter (Header 1500×500, Timeline Image 1200×675).
Will resizing a PNG preserve transparency?+
Yes. The Canvas context is configured with alpha support. Transparent PNG logos and icons are resized without adding any solid background fill.
What happens if I scale up a small image?+
Upscaling always introduces some softness or pixelation — no algorithm can invent detail that wasn't in the source. Lanczos-3 minimises ringing artifacts compared to other upscale methods, but for significant enlargements (2×+), starting with a higher-resolution source image will always produce better results.