Home / Images / AI upscaler

AI upscaler

Upload an image, choose a resolution (2K ¥2, 4K ¥3) and a scene (portrait, landscape, architecture and 53 more, or leave it on the default prompt); after payment succeeds AI repaints it sharper at the original aspect ratio and you can download the result. Failed runs are refunded automatically.

Pay per run (2K ¥2, 4K ¥3), no sign-up, Alipay online payment only. The original image is sent to our backend for AI inference only after payment succeeds; nothing is written to server storage, selection and coordinate mapping happen in your browser, and failed runs are refunded automatically.

More tools in Images

Frequently asked questions about AI upscaler

How is this different from AI image generation?
They work in opposite directions: image generation writes a description and draws a brand-new image from scratch, while upscaling takes an existing image and repaints it more clearly. So this page needs no description: just upload an image and choose a resolution, and the ratio always follows the original instead of being cropped to a square.
How much does this tool cost?
It is charged per use in two tiers: standard 2K ¥2 per run (area about 2048×2048) and high definition 4K ¥3 per run (area up to 4096×4096, four times 2K). Clicking "Start processing" opens the payment window first; pay for the tier currently selected and processing starts automatically once the payment succeeds, with no second click. One payment covers exactly one image, and a failed run is refunded automatically to the original payment method. Switching tiers requires a new payment, and a paid order that goes unused is refunded automatically.
Should I choose a scene? What does it do?
It is optional. If you choose a scene (56 categories such as people, landscape, architecture, cars and food), the system switches to a more targeted prompt, for example portraits leaning towards skin texture and hair, or screenshots of text leaning towards small-type legibility; leave it unset and the default prompt is used. The scene affects neither the price nor the output size.
Can you guarantee it stays the same image? Might it be altered?
The prompt says explicitly that content, composition, ratio, text and colours must stay unchanged and that nothing may be added, removed or moved, but generative repainting is still stochastic: facial features, small text and dense textures may be filled in slightly differently from the original, and the actual result is what counts. Keep the original for important material.
Is the image cropped to a square on output?
No. The backend scales the output up proportionally from the original ratio to the pixel area of the chosen tier (then aligns it to a multiple of 16), so a portrait stays a portrait and a wide image stays wide. The page shows the expected output before payment, and the result card carries a "ratio unchanged" line so you can check directly.
Are images uploaded to a server?
Yes, but the image is submitted for processing only after payment. So that the model can fetch the image, it is uploaded once after you pick it; the page keeps nothing but an image reference and never shows or hands out an image link. The result comes back to the browser as base64 and is not written to disk. Please do not upload sensitive content such as ID documents or private photos.
How long does processing one image take?
Commonly a dozen or so seconds; 4K takes a little longer. The page shows the status live; on timeout you get a notice and an automatic refund, so you can retry directly. Only one image is processed at a time: clicking again while the previous one is still processing shows "try again later", and that attempt is not charged.
Will processing overwrite the original file?
No. The original stays read-only and the result is saved as a new download you can review before replacing anything; the original file is never touched.
Which image formats are supported?
Input accepts any common format the browser can decode (JPEG, PNG, WebP, GIF, BMP and more); export supports JPEG, WebP and PNG, and each tool offers whichever of those its page lists (code-to-image exports PNG or JPEG, image-to-ASCII exports TXT / PNG / SVG / HTML).
Are images uploaded to a server?
Not for anything except AI watermark removal, AI background removal, AI image generation and AI upscaling. Compression, cropping, collage, image-to-Base64, code-to-image and image-to-ASCII all run locally in the browser, so images never leave your device; the four AI tools need model inference, so the content (the image or the description) is submitted to our backend only after payment succeeds, and the result image is sent straight back to the browser. The source image used by AI image generation and AI upscaling is also uploaded to our site, only so the model can fetch it; the page never shows or hands out an image link, and the result image is still never written to disk.

About Local Toolbox

Local Toolbox gathers the small tools you reach for every day into one place: formatters, encoding converters, date maths, ID card and address parsing, image and QR code handling, password and amount conversion. Open a page and start working — no sign-up, and almost every tool has no usage limit.

Most features compute inside your browser, so your input never leaves your device; the pages are plain static assets, load fast and keep working offline. The IP page asks our backend for your outbound address when it opens and sends none of your input. AI watermark removal, AI background removal, AI image generation and AI upscaling send content to our backend for inference (watermark removal is ¥1 per run; background removal is ¥1 per run; image generation and upscaling are priced by resolution — 2K ¥2, 4K ¥3); AI annotation sends images to a large model that locates the target and returns coordinates (¥0.5 per image, refunded only when every image in the batch fails); AI flowchart, AI mind-map and AI Mermaid chart generation send your text description to a model that builds the diagram (three separate products, charged ¥0.5, ¥0.5 and ¥0.5 per run respectively); and the accounting helper sends the income, expense, social-insurance and balance figures you fill in to generate a report (¥0.5 for the spoken analysis, ¥1 for the report). These features accept Alipay only, upload nothing before payment succeeds, and refund automatically if processing fails.

Links