How to use ImgScale Studio

Updated 2026 · ~6 min read · by kizura

ImgScale Studio turns a series of single-purpose tools into one repeatable flow. Instead of upscaling an image, downloading it, re-uploading it to remove the background, downloading again, and so on, you build the sequence once and run any image — or a whole folder of them — through every step in a single pass. This guide walks through it from scratch.

What Studio is

Think of Studio as a conveyor belt for images. Each station on the belt is a tool — upscale, remove background, resize, crop, compress, convert, watermark. You arrange the stations in order, drop images onto the belt, and each one comes out the other end fully processed. Like every ImgScale tool, the whole pipeline runs in your browser — nothing is uploaded.

Step 1 — Open Studio

Go to imgscale.net/studio. There's nothing to install and no account to create. The page loads with an empty pipeline and a panel of available actions. On first use, tools that rely on an AI model (like upscaling or background removal) download that model once and cache it locally; after that they run offline.

Step 2 — Add pipeline steps

Each action you can chain appears as a node you add to the pipeline. Add a step by selecting it from the actions panel — it drops into the flow as a new node. Add as many as you need; a typical pipeline has two to five steps. Available steps include:

• Upscale — enlarge and sharpen up to 4× with AI
• Remove background — cut out the subject automatically
• Resize — scale to exact pixels or a social preset
• Crop — trim to a ratio or region
• Compress — shrink the file size
• Convert — change format between PNG, JPG, and WEBP
• Watermark — stamp text onto the image

Step 3 — Configure each step

Most steps have options. Click a step to open its settings: set the upscale factor (2× or 4×), the exact resize width and height or a preset, the compression target, or the output format for a convert step. Each node remembers its own settings, so you can, for example, resize to one size early in the pipeline and compress to a specific quality at the end.

Step 4 — Order the steps correctly

Order matters, because each step receives the output of the one before it. A few rules of thumb that produce the best results:

• Upscale early. If you're enlarging, do it first so later steps work on the high-resolution version.
• Remove background before resizing or cropping so the cutout is computed on the full subject.
• Crop and resize in the middle to set the final dimensions.
• Compress and convert last. File-size and format changes should be the final stations, after the pixels are settled, so nothing re-expands the file afterward.

A clean general order is: upscale → remove background → crop → resize → convert → compress.

Step 5 — Drop images & run

With the pipeline built, drop in your image (or images) and run. Each file is fed through every step in order and the finished results are offered together for download. Because all of it happens locally, there's no upload wait and no per-step round trip — the whole chain executes in one go on your device.

Batch processing

Studio's biggest time-saver is batch mode: drop in many images at once and every one passes through the identical pipeline. This is ideal when the same edit applies to a whole set — a store owner preparing dozens of product photos, a creator resizing and compressing a week's worth of posts, or anyone standardizing a folder of images to one size and format. Build the flow once, drop the batch, and download them all processed the same way.

The ⌘K command palette

Press ⌘K (or Ctrl+K on Windows) to open the command palette — a fast keyboard-driven way to add steps and jump around without reaching for the mouse. Start typing an action name (“resize”, “compress”) and select it to add it to the pipeline. For anyone building pipelines often, it's the quickest way to work.

Ready-made pipeline recipes

Three pipelines that cover the most common needs:

Product photo for a store

Remove background → place on white → resize to a square (e.g. 1000×1000) → compress. Produces clean, uniform catalog images that load fast.

Social media post

Upscale a small source → crop to the target aspect ratio (1:1 or 16:9) → add a watermark. Turns a low-resolution image into a polished, branded post.

Web optimization

Resize to display width → convert to WEBP → compress. The standard recipe for fast-loading site images.

Tips & troubleshooting

• Start small. Test a pipeline on one image before running a large batch, so you can confirm the order and settings.
• Large batches take time. Everything runs on your hardware; a big batch through an AI step (upscale, remove background) is the slowest part. Mid-range laptops handle it fine, just not instantly.
• Compress last. If your output files seem larger than expected, check that compression is the final step — a resize or convert after it can undo the savings.
• Nothing uploads. Open your browser's Network tab and watch it stay empty while a batch runs — confirmation that your images never leave your device.

Ready to build one? Open ImgScale Studio and drop in your first image. Related: AI upscaler · Remove background · Compress · Main guide

Three ready-made pipeline recipes

Once the node canvas makes sense, most real work settles into a few repeatable patterns. These three cover the requests we hear most, and each takes under a minute to assemble.

• The listing-photo pipeline — Resize → Compress: drop a folder of product shots, resize to 1,600 px, compress to JPG 80. Every marketplace gets consistent dimensions and sub-1 MB files, and the batch comes back as one ZIP instead of thirty downloads.
• The email batch — Resize 1,600 → Compress q75: the recipe from our email guide, automated. A phone-photo dump becomes attachable in one run; in our measurements this pair of nodes cuts a typical detailed photo by roughly three-quarters.
• The archive-prep pipeline — Grayscale → Resize → Compress: for scanned documents and receipts where colour is noise, grayscale first (it also compresses better), then normalise the size. Consistent, searchable-looking archives without opening an editor.

Two habits make pipelines pleasant: order nodes from “changes pixels most” to “changes pixels least” (geometry before compression, always — compressing and then resizing re-encodes twice), and when experimenting, test on one image before pointing a node chain at two hundred. Smart Batch exists for the opposite case: when a mixed folder needs different treatment per file, let it inspect and recommend, then override the odd ones out by hand.