Live
Image
Resize Image Online
Set exact pixel dimensions, scale by percentage, or fit to a bounding box — free, no account.
Open toolHome › Tools › Image Processing › Enlarge Image
Enlarge Image Online — Free
Scale up to 4× with Lanczos quality filtering. No AI required — no account, no watermark.
Drop an image here
or click to browse · max 10 MB · JPG, PNG, WEBP, AVIF, BMP, TIFF, HEIC
()
Scale factor
Aspect ratio is preserved automatically.
Enter a width between 1 and 10000 px.
Frequently asked questions
Is the image enlarger free?
Yes — completely free with no account required. No watermarks are added to your enlarged files, and no subscription is needed. There is a 10 MB upload limit per file.
What is Lanczos upscaling and how does it work?
Lanczos is a high-quality resampling algorithm that uses a windowed sinc function to calculate new pixel values when enlarging an image. It preserves sharp edges and fine detail better than simpler methods like bilinear or nearest-neighbour scaling. It is the best general-purpose algorithm for upscaling photos and graphics without AI processing. For 2× enlargement of clean photos it produces excellent results; for very degraded or low-resolution images, AI upscaling tools (Canva, Adobe Firefly) may give sharper results.
What image formats can I enlarge?
The enlarger accepts JPG, PNG, WEBP, AVIF, BMP, TIFF, and HEIC files up to 10 MB. You can download the result in the same format as the original, or switch to JPG, PNG, or WEBP on output.
Keep going
More image tools
Enlarging an image increases its pixel dimensions — turning a 1000 × 750 photo into 2000 × 1500 or 4000 × 3000. The challenge is that upscaling cannot add detail that was not in the original; it can only interpolate from the pixels that exist. The quality of that interpolation determines whether the enlarged image looks sharp or blurry.
AT USE Image Enlarger uses Lanczos resampling — a mathematical filter that applies a windowed sinc function across a three-lobe neighbourhood — which produces sharper edges and finer detail retention than the bilinear and bicubic algorithms used in basic browser-based scaling. Unlike nearest-neighbor interpolation, which produces blocky pixel artifacts, Lanczos distributes the calculation across a wider sample window, reducing ringing artifacts while preserving edge contrast. It is the same algorithm used in professional image editors and video production pipelines when quality matters more than encoding speed.
Scale options
The tool offers four multiplier presets — 1.5×, 2×, 3×, and 4× — alongside a custom target width field. Multipliers scale both dimensions proportionally: a 2× enlargement of an 800 × 600 image produces exactly 1600 × 1200 pixels. The custom width field accepts a pixel value and calculates the corresponding height to maintain the original aspect ratio precisely. There is no option to change width and height independently — if you need a different aspect ratio, crop the image first, then enlarge.
When enlargement works and when it does not
Lanczos produces good results on photographic content with smooth gradients and natural textures, and acceptable results on simple graphics with clear lines and solid fills. It does not produce good results on small screenshots with anti-aliased text: there is not enough source detail to represent the characters sharply at a much larger size, so they will appear blurry regardless of the algorithm used. For text-heavy screenshots, the only reliable path to a larger image is recapturing the original at higher resolution.
File size limit
The per-file cap is 10 MB (not the 20 MB limit on most other tools on this site). A 4× enlargement multiplies the pixel area by 16, which can produce very large output files from large inputs. The 10 MB cap keeps processing within predictable memory bounds on the server. If your source file exceeds 10 MB, compress it first with the Compress Image tool, then enlarge.
Common use cases
- Print preparation from web photos — Web images are typically captured or exported at 72–96 DPI. Doubling the pixel count raises the effective print resolution at the same physical size. A 900 × 600 photo scaled 2× to 1800 × 1200 prints at the same dimensions with twice the pixel density — noticeably sharper on most desktop printers.
- Fitting a template minimum size — Slide decks, banner templates, and social post templates often specify a minimum width (for example, 1200 px). A photo that is only 600 px wide will appear soft when stretched by the template software to fill that dimension. Enlarging it before inserting gives the template a sharp source to work with.
- Documentation and knowledge-base screenshots — Technical writers including small UI screenshots — icons, badge images, cropped interface details — in documentation often need them 2–3× larger to be readable on high-density displays without hunting for a higher-resolution original.
- Social profile and header images — Platform-specified minimum pixel requirements for profile photos and headers (LinkedIn cover photo: 1584 × 396 px; GitHub profile: 460 × 460 px) are often larger than an existing headshot or logo. Enlarging to meet the minimum before cropping is faster than reshooting.
- Archival scan upscaling — Document scans done at 150 DPI or below produce low-resolution image files. Scaling 2× gives the archivist a larger working file before any sharpening pass in a desktop editor, without generating the AI-invented texture that neural upscalers add.
Frequently asked questions
Will enlarging fix a blurry photo?
No. Lanczos resampling preserves and refines the pixel relationships already present; it cannot reconstruct detail that was blurry or out of focus in the original. If the source image is blurry, the enlarged version will be a larger blurry image.
What is the difference between Lanczos and AI upscaling?
AI upscaling tools use machine-learning models trained on large image datasets to generate plausible-looking detail at larger sizes. Lanczos is a deterministic mathematical filter — no detail is invented, but edges stay sharper and artifacts are lower than simpler filters. AI upscaling is better suited to photographs where synthesised texture is acceptable; Lanczos is better for graphics, logos, and technical diagrams where accuracy matters over appearance.
Why is the file size limit 10 MB instead of 20 MB?
A 4× enlargement multiplies pixel area by 16. A 10 MB JPEG input at 4× can produce output exceeding 80 MB in memory, which causes unreliable processing on the server. The 10 MB cap keeps output within predictable bounds. Compress the source file first if it exceeds 10 MB.
Can I enlarge and convert the format in the same step?
Yes. The output format selector is independent of the input. Upload a HEIC from an iPhone, set 2×, and download a PNG — format conversion and enlargement happen in the same operation.
Does enlarging preserve transparency?
Yes. Transparent pixels in PNG and WEBP source files are enlarged with the same Lanczos algorithm as colour channels. Choose PNG or WEBP as the output format to retain the alpha channel in the result.
Is this free?
Yes. No account required, no watermark on the output, no usage cap beyond the 10 MB per-file technical limit.