Google Photos launches AI-powered face retouching: Here's how Touch Up and the new smart editor work

  • Google Photos incorporates Touch Up, an AI-powered facial retouching system that allows you to edit up to six faces separately in the same photo.
  • The machine learning models run locally (package of about 16 MB) and offer intensity controls for skin, dark circles, iris, teeth, eyebrows and lips.
  • The new features are integrated into the revamped Google Photos editor alongside Reimagine, Magic Editor, AI Enhance and Remix, with support for C2PA Content Credentials.
  • The rollout begins on Android with minimal RAM and system version requirements, and reopens the debate on mental health, beauty standards, and transparency in edited images.

AI-powered facial retouching in Google Photos

Google has taken another step in the integration of the Artificial intelligence within the Google Photos app With the launch of new facial retouching tools that work directly from the image editor, the company is reinforcing its commitment to transforming the photo gallery into a true mobile editing studio, without the need for external applications.

This new package, known as Touch Up inside Google PhotosIt allows users to individually adjust specific facial features in a photo, from skin texture to iris brightness and tooth color. The update is being rolled out gradually to Android devices, with minimal memory and system version requirements, and is part of a broader overhaul of the editor based on Gemini and other advanced AI features.

What is Touch Up and what can it do for your photos?

Touch Up is the name of the new set of facial correction tools based on AI models which Google has incorporated into its Google Photos editor. Far from being a simple beauty filter that acts on the entire image, it is a system designed to work face by face and feature by feature.

In practice, these retouching options focus on photographs where faces are detectableOnce the image is open, the user can select a specific face and access a menu with various specialized tools. The logic is simple: you choose the area to enhance, adjust the intensity level with a slider, and see the result in real time, without having to struggle with complex editing parameters.

Google positions Touch Up as a way to apply quick and subtle changes Before sharing a photo, these tools are designed to correct minor details without completely altering the person's appearance. However, the actual impact will depend on how far each control is adjusted, something that is entirely up to the user.

The company insists that the goal is for the images to better reflect "how you feel at that moment," but in practice we are facing a granular face editing tool accessible to anyone with a compatible mobile phone, bringing these features to a very wide audience, including in Europe and Spain, where Google Photos is one of the most used gallery apps.

Available tools: from dark circles to teeth whitening

Google Photos' new face editor groups its functions into a series of Specific adjustments for different areas of the faceIn the interface (located in the "Actions" or "Tools" menus depending on the version), options such as the following appear:

  • Smooth skin: reduces the appearance of pores, rough textures, or small skin imperfections.
  • Under-eye correctionIt brightens the area under the eyes and reduces shadows that give a tired appearance.
  • Iris enhancement (irises): increases the brightness and contrast of the eye, making it look more "alive" in the photo.
  • Teeth whiteningIt lightens the tone of the teeth without needing to modify the rest of the face.
  • Eyebrow shaping: corrects or highlights the shape and presence of the eyebrows.
  • Lip modification: allows you to subtly enhance the contour and color of your lips.

Each of these tools is managed with its own independent intensity controlThis means that you can apply, for example, a very light skin smoothing to one person and a more noticeable teeth whitening to another in the same photo, without a single filter changing everything equally.

This separation by features is what brings the Google Photos experience closer to that of professional retouching applications rather than simple social filtersWhere many services like Instagram or Snapchat offer a closed effect with a single global slider, Touch Up breaks down the settings more finely, resulting in less generic results when used in moderation.

In group photos, this logic is especially useful: the function It automatically detects up to six faces in a single image. And it allows you to edit each one separately. In practice, you can correct one person's dark circles without changing the expression of the person next to them, something that until now was more typical of advanced desktop editors than a pre-installed mobile app.

Face editing: up to six faces in the same image

One of the most striking new features of Touch Up is its ability to to work individually on each face within a group photographThe system analyzes the image, identifies faces, and highlights them as separate, editable elements.

From the app, the user can tap each face to open the facial tools panel and decide what kind of retouching to apply. If the photo contains more than six faces, Google Photos displays a warning that the feature's limit has been reached and only allows working with the first six detected, which limits its use for portraits from large events or photos of very large groups of people.

This individualized approach makes a clear difference compared to apps based on very broad generative filters, like some FaceApp modes or tools that apply the same style to all subjects in the scene. While those options usually alter the entire image at once, here the editing is adjusted face by face and feature by feature.

For those who manage corporate or social media content in Spain or the rest of Europe, this precision helps to obtain Studio-quality results without leaving the galleryespecially in team photos, portraits for company websites, or images for presentations where these details are usually given more attention.

At the same time, the tool is not positioned as a system for radically transforming the face —there are no changes in age, gender or extreme alterations— but as a layer of moderate correction, geared towards small cosmetic touch-ups for everyday use.

Local processing and technical requirements in Android

Behind these functions lies a set of machine learning models that are distributed and installed on the device itselfAccording to documentation shared by Google, Touch Up downloads a package of about 16 MB the first time it is used, which is stored for later edits.

The consequence of this design is that Most of the processing is done locally on the phone, without needing to send the photo to the cloud to apply the adjustments. This has important implications for both performance and privacy: it reduces the waiting times associated with sending data and, in theory, also the exposure of biometric information to external servers.

Regarding compatibility, Google states minimum hardware and system requirements for the editor to run smoothly. In the most advanced version of Touch Up, the company mentions devices with at least 4 GB of RAM and Android 9.0 or higher, while other sources mention a more basic activation starting with Android 8.0 and 3 GB in certain markets or models.

This range indicates that the functions have been designed to relatively recent devices, but not necessarily high-endThis is particularly relevant in Europe, where several generations of models coexist. Older or entry-level phones may experience some slowness or lack access to all options, especially regarding more demanding effects or integration with other AI features of the editor.

For now, Google is focusing its rollout on Android and hasn't provided a firm timeline for iOS. In any case, this type of release typically involves a phased deployment by region and modelTherefore, it is possible that not all users will see the new tools at the same time in Spain or other European countries.

A revamped photo editor: Reimagine, Magic Editor, AI Enhance, and Remix

The facial retouching functions come with a broader reorganization of the Google Photos editor, which in 2026 has received several improvements aimed at simplifying access to existing AI tools.

The menu now groups options such as Reimagine, Magic Editor, AI Enhance and RemixThese features allow you to do everything from transforming a scene's background to turning an everyday image into a cinematic version. The goal is to clean up the interface, reduce the number of submenus, and make the available features more visible.

Reimagine stands out for offering a text-guided editing in specific areas of the photoThe user can select an area and describe what they want to happen, for example, completely changing the sky, adding storm clouds, or redesigning the background to resemble a different location. The model then generates the new content in a way that is consistent with the rest of the scene.

Magic Editor, for its part, focuses on more classic assisted editing actions, such as changing the framing, removing unwanted objects, or repositioning elements within the composition with results that previously required considerable mastery of desktop programs.

AI Enhance acts as a button Overall improvement in lighting, color, and detailwith one or two intensity levels for those who prefer not to delve into manual adjustments. And Remix takes a more creative approach, proposing artistic styles and visual reinterpretations that can transform a travel photo into something closer to a film still.

Gemini and conversational image editing

Another fundamental change in the Google Photos ecosystem is the increasing integration of Gemini as a natural language-based editing engineInstead of navigating through a menu tree, the user can describe in words what they want to change and let the system take care of it.

This translates into orders of the type "Remove the dog from the background" or "Make the sky more dramatic"The model interprets this to apply a combination of tools: object selection, generative fill, color adjustments, etc. The idea is that the technical barrier to editing will almost completely disappear.

In this context, new facial retouching tools indirectly benefit from the AI infrastructure already deployed in the Google ecosystemAlthough for now they maintain a more classic interface based on sliders, it wouldn't be surprising if, in the medium term, they also add conversational editing: instructions like "remove my dark circles a little" or "whiten my teeth a little" fit perfectly with this approach.

For European users, where regulations and public debate on the Generative AI and the processing of personal data They are gaining ground; this type of integration reinforces the idea that conversational assistants will play an increasingly central role even in seemingly simple tasks such as retouching a photo before uploading it to social media.

Content Credentials, SynthID and transparency in retouched photos

The expansion of editing capabilities is accompanied by an additional effort on the part of Google in the area of transparency and authenticity of the imagesThe company has begun integrating support for C2PA Content Credentials into Google Photos, a standard promoted by the Coalition for Content Provenance and Authenticity.

These credentials add a set of verifiable metadata that indicates how and when it was editedIn theory, anyone who receives the image can check that information, both on compatible devices and through verification tools, and find out if AI retouching has been applied, including changes made with Touch Up.

The C2PA rollout initially begins in Pixel 10 devices and will be extended to other models over time. Meanwhile, Google was already using invisible SynthID watermarks in images generated with tools like Reimagine, a technology that allows identifying content created or modified with AI even when screenshots or cropping have been done.

For users in Spain and the rest of Europe, where specific regulations are discussed regarding identification of synthetic content and deepfakesThe combination of visible Content Credentials and invisible watermarks represents an intermediate step: it does not prevent the photo from circulating, but introduces the possibility of tracking and verifying whether it has undergone an automated editing process.

The challenge lies in how many people will actually consult this metadata and how these signals will be integrated into platforms such as social networks, media outlets, or messaging tools, areas where retouched images often circulate without additional context.

Impact on mental health, beauty standards, and image rights

The arrival of such accessible facial retouching tools did not occur in a social vacuum. Various studies and analyses, cited by international technology media, indicate that the intensive use of filters and retouching in personal portraits It can contribute to problems with self-esteem, a distorted perception of one's own body, and pressure to conform to certain aesthetic standards.

With features like Touch Up built into an app used daily to view and share photos, the line between a harmless, one-off editing and a daily practice of systematically altering appearance It becomes more difficult to track. Especially among teenagers and young adults, who are the ones who interact most with visual content on social media.

In addition to this, there is another less discussed but equally relevant issue: the image rights in group photosIf someone retouches another person's face in a team photo, event, or meeting without asking permission, they are not just applying a technical adjustment; they are altering how that person appears to others.

In European work or educational environments, where there is a growing sensitivity towards the management of public image and consent, this practice may clash with internal policies or even regulatory frameworks on data protection and the use of photographs.

Content Credentials provide a degree of traceability, since They state that the photo has been edited.But they do not solve the problem of consent on their own: it is still up to users and organizations to decide when and how it is acceptable to apply these retouching techniques, and to communicate this clearly if the images are used in professional or commercial contexts.

Practical use: from family albums to professional content

Beyond the ethical debates, in everyday life there are countless situations where these functions can be practical. For the average user, the ability to to soften dark circles in a family photo, correct an odd shine on the teeth, or to minimize a single pimple Before printing an album or sharing the image with friends, it may be enough to justify its use.

In the professional sphere, especially in small businesses, freelancers or startups in Spain and other European countries, having a facial retouching tool integrated into the gallery itself It reduces reliance on paid services or external designers for simple tasks. For example, preparing team portraits for the "About Us" page or refining photos for a press kit without leaving Google Photos.

By working natively on images already stored in the Google cloud, Touch Up integrates seamlessly with the rest of the workflow: capture with the mobile camera, automatic upload, quick editing and direct sharing on social networks, messaging or email.

However, the fact that the tools are convenient and free does not eliminate the need for define certain internal limitsIn organizations that frequently publish images of real people—customers, employees, students, patients—it may be advisable to establish basic policies on what types of retouching are allowed and which are not, to prevent everyday use from resulting in an image too far removed from reality.

In short, Google Photos has gone from being a simple gallery app to becoming something else in just a few years. an advanced AI-powered photo editorWith Touch Up, this leap is consolidated in the field of portraiture, bringing to mobile phones functions that were previously associated with professional software and, at the same time, opening up new questions about how we want to see ourselves—and show ourselves—in our daily photos.


Add as preferred source