Topaz Photo AI

Overview

Photo AI folds the separate quality problems into one application and, more importantly, into one decision. It analyses an image, works out whether it is noisy, soft, too small, badly focused or all four, applies the models it judges appropriate at the strengths it judges appropriate, and shows you the result. If the automatic judgement is wrong you can override every part of it, but for a large batch of ordinary photographs it usually is not.

Denoise handles sensor noise and high sensitivity grain without turning skin into plastic or fine texture into smears, which is where conventional noise reduction fails. The models distinguish genuine detail from noise rather than blurring everything below a threshold, and separate models are trained for raw files and for images that have already been processed and compressed.

Sharpening is a reconstruction rather than an edge contrast trick. Separate models handle general softness, motion blur and missed focus, each attempting to reverse the specific way detail was lost rather than simply increasing local contrast around edges. Combined with the enlargement models it means a small, soft, noisy image can come out as something usable at print size.

Face recovery is the feature that decides most portrait work. Faces are detected individually, each one can be included or excluded, and the strength is adjustable per face so a sharp foreground subject is left alone while a soft background face is rebuilt. Raw files can be processed before demosaicing for the best results, output writes to TIFF, JPEG, PNG and DNG with profiles preserved, batch processing handles a folder unattended, and the application installs as a plugin into the common photo editors.

Key Features

• Automatic analysis choosing which corrections an image needs and at what strength
• Denoise models trained separately for raw files and processed compressed images
• Sharpening models for general softness, motion blur and missed focus
• Enlargement models for increasing resolution while reconstructing detail
• Per face detection with individual inclusion and adjustable recovery strength
• Raw processing before demosaicing for the highest quality result
• Full manual override of every automatic decision the analysis makes
• Batch processing across a folder with consistent or per image settings
• Output to TIFF, JPEG, PNG and DNG with colour profiles preserved
• Plugin integration into the common photo editors and catalog applications

What Changed in This Build

• Improved automatic strength selection on mixed quality batches
• Better skin texture retention at high denoise strengths
• Faster processing on current generation GPUs
• Face detection reliability improved on group photographs

System Requirements

Processor	Intel or AMD 64-bit, 4 cores or more recommended
Memory	16 GB recommended
Graphics	GPU with 6 GB VRAM recommended, CPU processing available without one
Storage	12 GB for install with all models
Display	1920 x 1080 or higher

Release Details

Full title	Topaz Photo AI
Version string	4.1.0
Publisher	Topaz Labs
Category	Photo Editing
Licence	Full version, all models included
Platform	Windows 10 and 11, 64-bit
Interface language	English
Archive size	2.48 GB
Mirrors	4 active, all reporting online
Listed	refreshed 3 months ago, checked again 40 minutes ago

Installation Notes

• Check the automatic settings on a few images before running a large batch. The analysis is good but it is not infallible on unusual material.
• Process raw files here before your normal editing pass rather than after, since it works best on data that has not been compressed yet.
• Turn face recovery down rather than off on portraits. At maximum strength it can start inventing features that were never in the frame.

Before You Start

Read the notes above before running setup. Most of the problems people write in about are covered there, and almost all of them come down to an older version of the same software still being installed, a graphics driver that predates the release, or a security suite quarantining part of the package mid install.

There is no archive password on this or any other entry in the library. If something you downloaded elsewhere claims to come from here and asks for one, it did not come from here. Nothing on this site is behind a survey, a shortener or a countdown either.

If the entry does not match what the page describes, say so on the contact page and include the version string from the release details table. Reports naming a specific version get checked first because the checker can reproduce them straight away.

Frequently Asked