Download Topaz Gigapixel AI for Mac

In the edit

You’ve got a folder of old family portraits scanned at 72 dpi, or maybe a client sent over a web-sized JPEG they want printed. Normal upscaling turns them into blurry messes, but that’s exactly the problem Topaz Gigapixel AI is built to solve.

• Drag a set of portraits into the app – it handles TIFF, JPEG and PNG right out of the box, and the batch queue lines them up without fuss.
• Pick a preset for face recovery. The AI model restores detail in eyes and skin texture that standard interpolation would lose completely.
• Set the scale to 4x for prints. The preview updates in real time, so I can zoom in and check that the wrinkles and hair strands actually look natural.
• Hit process and let the batch run while I edit something else.
• Export as 16-bit TIFF to keep the extra detail for further colour work in Capture One or Photoshop.

The whole workflow stays inside one window, no switching between tools. For anyone who regularly works with old photos or web-sourced images, Gigapixel AI turns a frustrating step into something you can trust and walk away from.

Dragging in a batch of 1970s newspaper scans

I usually start Gigapixel AI when I’ve got a batch of images that need rescuing — maybe a dozen old newspaper scans from the 1970s that a client wants to reprint for a family history project. I drag the whole folder straight into the app’s window. The thumbnails appear in a grid, and the batch queue shows how many files are waiting. No need to open them one by one.

From there I pick a single image to set the parameters. The preview panel shows the original on the left and the upscaled version on the right, and I can drag the divider to compare specific areas. I usually zoom into a face or a section of fine text to see what the AI is doing. The model options sit in a panel on the side — I tend to start with Standard for general photos and switch to Face Recovery if the subject is a person. The scale slider goes from 2x up to 6x, but for most print work 4x is enough.

Once I’m happy with the preview, I apply the same settings to the whole batch. The processing runs in the background, and the app shows a progress bar for each image. On a busy day I’ll start the batch, then switch to Lightroom to cull another shoot. When I come back, the upscaled files are sitting in the output folder I chose at the start.

The final step is export. I save as TIFF at 16-bit depth because I know I’ll be doing colour grading later. If the images are going straight to a client who only needs screen resolution, I export as JPEG at 8-bit. The app remembers the last output folder, so the second batch is even quicker.

Reviving 6-megapixel DSLR photos from the mid-2000s

Photographers who shoot with older digital cameras benefit a lot. Those 6-megapixel DSLR files from the mid-2000s hold up surprisingly well after a 2x or 3x upscale — the AI reconstructs detail that was always there but too small to print large. I’ve had friends who shoot weddings on a Canon 5D classic get genuinely usable 24‑inch prints from Gigapixel AI.

Designers and artists who work with sourced imagery also find it useful. If you’re pulling reference photos from the web or extracting stills from video, the resolution is almost never high enough for print. Running those files through the app before placing them in an InDesign layout saves a lot of manual retouching.

Archivists and family historians who scan old prints or documents at moderate resolution can double or triple the usable size without losing the character of the original. The AI doesn’t invent detail where there’s none — it just fills in what the optics and sensor should have captured. That distinction matters when you’re handling historical material.

Placed between Capture One and Photoshop

Gigapixel AI sits in my workflow between the raw processor and the editor. I’ll do basic exposure and white balance in Capture One, then send the TIFF to Gigapixel for upscaling, and finally open the result in Photoshop for dust spotting and final sharpening. The app doesn’t try to replace those tools — it just does one thing well and passes the file along.

On macOS, I use Automator to set up a simple folder action. Any image I drop into a designated folder gets processed automatically with the last settings I used. That’s useful when I’m working through a large archive and want to keep my hands on the mouse for sorting, not clicking through dialogs.

The app also respects the standard macOS file dialogs and drag-and-drop behaviour, so it feels native. I can drag images from Finder straight into the queue, and the output folder is accessible from the sidebar. No surprises.

Restoring a 1960s wedding album to 12x18 prints

A friend of mine runs a small restoration service for old wedding albums. She recently got a set of scans from a 1960s album — the originals were small prints, scanned at 300 dpi, but the client wanted 12x18 inch enlargements for a wall display. That’s roughly a 4x upscale from the scanned resolution.

She loaded the 40 images into Gigapixel AI, applied the Face Recovery model to the first one, checked the preview on a few close-ups of the bride and groom, then applied the same settings to the whole batch. The processing took a while, but she left it running overnight. Next morning, the files were ready. She opened a few in Photoshop to adjust the yellowing from the old paper stock, then delivered the set as high-resolution TIFFs.

The client couldn’t tell the images had been upscaled. That’s the kind of result that keeps people coming back.

Mac it needs

• macOS 12 or later
• Intel or Apple Silicon Mac
• Interface language: English