The past couple of years have been a wild ride in the tech world. Namely because AI became such a huge buzzword that everyone wants on that particular bandwagon. Google announced a few different AI projects — from Bard to DeepMind to Gemini (it’s all unified under the Gemini name now, thankfully), Samsung released its
.
With all of these manufacturers and developers throwing AI features at us, most of which cover the same ground, it quickly becomes a confusing mess. Which one does what, who has exclusive features, who does it best?
Let’s take a step back, take a deep breath, and look at all the currently known features for Apple’s AI, Google’s Gemini, and Galaxy AI.
We will be including the promised features as well — since Apple is still in the middle of rolling out the complete Apple Intelligence feature set. We figured it’s better to throw them in as well, so we can have a clearer picture of which platform is going for what, even if it’s not quite there yet.
Both the Google Pixel and Galaxy phones offer live translation on a call between a few of the bigger languages. It’s not flawless, but can be helpful in certain scenarios.
Smart reply
Apple AI – Yes | Gemini – Yes | Galaxy AI – Partial
The language models can pretty accurately interpret what a message you have received is about, and then generate a more elaborate reply for you. Apple AI and Gemini do this in email and text, Galaxy AI currently only does if in the Messages app (texts).
Compose and rewrite text
Apple AI – Yes | Gemini – Yes | Galaxy AI – Yes
You can have the AI arrange your text neatly, compose entire paragraphs, or change the tone and style of a piece of text. Apple’s AI can also call up the help of ChatGPT for longer, more elaborate, essay pieces. In any case, all three assistants here cover this feature.
Image-related AI features
Generate images
Apple AI – Yes | Gemini – Yes | Galaxy AI – Partial
Apple has its own Image Playground, which generates animation-style images that generally aren’t… very diverse or usable. But, again, with ChatGPT baked into iOS 18, you can just ask it to generate more realistic images, or ones in different styles.
It’s worth noting that Apple also has the new “Genmoji”, which will specifically generate emoji-style images based on your image prompts. These are… met with lukewarm reception, as they kind of defeat the purpose of expressing yourself with emoji — the fun is in their limited numbers and trying to create different combinations and meanings.
Google’s Pixel Studio is also a bit better at making images out of thin air, but it’s still a bit limited — specifically, it will refuse to do human figures. Google also has the Emoji Lab, which will mix the characteristics of two emoji into one to create a sort of Franken-moji. These are a bit more fun than Genmoji.
Samsung’s AI can currently generate images from your own hand drawings, but doesn’t currently have the feature to make them from text prompts, nor emoji generators.
Visual search
Apple AI – Yes | Gemini – Yes | Galaxy AI – Yes
Apple AI has the new Visual Intelligence feature — press and hold the Camera Control button and then you can use the viewfinder to either ask ChatGPT or Google what you are looking at. Both Google and Samsung phones have Gemini with access to the camera, which works much like Visual Intelligence. There’s also Circle to Search, where you can perform an image search of anything you currently have on your screen.
Magic eraser
Apple AI – Yes | Gemini – Yes | Galaxy AI – Yes
This has been on Samsung and Google phones long before it was attributed to AI. Magic Eraser is a “generative eraser” tool, meaning it will delete any object from a photo and try to generate the missing background, based on context around it. Apple now has the Clean Up tool in Photos, which does much the same. One can argue that Apple’s tool is a bit more “AI”, since it will immediately identify and suggest which objects you may wish to delete from a photo.
Auto crop out subject from a photo
Apple AI – Yes | Gemini – No | Galaxy AI – Yes
Again, this has been around before the AI craze. Essentially, you can tap and hold on an object in a photo, and the software will “identify” its boundaries and allow it to “lift” it from the photo. Then, drag it to paste into another app, or generate a copy of that image with a transparent background. It’s a pretty useful tool for making YouTube thumbnails, and has been available on iPhone and Galaxies for a couple of years now.
Pixel exclusive: Photo Unblur, Reimagine, Add me, Magic Editor
Apple AI – no | Gemini – yes | Galaxy AI – no
Google’s been working on image-related tricks for some years now. The Photo Unblur feature will analyze and remove the blur from any picture in Photos. Even if you uploaded an old camera photo!
Magic Editor allows you to rotate and change the positioning of a photo. If you end up moving outside the lines of the original photo, Magic Editor will try to generate more background, based on the image.
Reimagine can change entire subjects in a photo based on a text prompt. Like “Make the grass yellow” or “Make the sky cloudy”.
Add me is a cool way to take a group picture without needing outside help. One person takes a photo of the group, then someone else steps in and directs the original photographer to take their spot in the frame. The AI will stitch picture 2 with picture 1 to make the entire group look as if they were together at the same time.
Some of these may come to other Android phones via a Photos update — Unblur is available, Magic Editor is in testing, and Reimagine may also be pushed at some point. But Add me, for example, is specifically a Pixel Camera feature.
General AI features
Contextual understanding and multi-layered commands
Google’s Gemini is the leader here. The conversational Gemini Live model is currently available, and in general, Gemini’s capabilities are comparable to ChatGPT’s. You can have Google’s assistant answer questions from the web, brainstorm or lead a conversation with you, or reach for limited information within apps and your account.
Apple’s Siri is supposedly getting a major overhaul this March, where it will also be able to follow conversations but also do complex, multi-layered tasks. Like pull information from one app to apply it to a query for another app. For the time being, we do have access to CharGPT, which is quite deep and elaborate, straight through Siri.
Bixby is currently lagging behind, but from what we know about One UI 7, which should be launching with the Galaxy S25, Samsung’s assistant should be getting a lot smarter. Apparently, it will be powered by Samsung’s own LLM model — Gauss — and Google’s Gemini. We’ll see how that pans out, but there are huge promises for a conversational style and multi-app operations.
Exclusive AI features right now
Apple Intelligence
Apple put a huge emphasis on re-doing how notifications are delivered with the help of AI. For one, messages that are more “important” will be automatically pushed to the top of your feed. Secondly, AI will be used to summarize long messages or chat strings, so you can get a better idea of what’s going on at a glance.
At least that’s the general idea. The execution is not quite there yet, with some summaries being pretty vague or funny.
The same treatment has been applied to the Email app, where a new Priority tab will be filtered and summarized by AI.
Google Gemini
Google did a couple of unique moves with its AI implementation. One is the new Screenshots app — any screenshots you ever take will be collected here, and AI image recognition will slap a few tags onto it. So, you can easily search for it later. Like “Phone number of plumbing service” or “Poster of a show with dates on it”.
The other unique feature is in none other than the Weather app. In the same spirit as summarizing a long article, Gemini will give you a quick recap of the weather forecast at the very top. Yeah, the level of usefulness here is arguable.
Then, we have a slew of other smart Google features, which were released long before everything had “AI” on it, but definitely fall under the umbrella:
- Call screening
- Direct my call
- Hold for me
- Clear Calling
- Reserve with Google
Samsung Galaxy AI
Samsung started off 2024 strong, with a slew of features in Galaxy AI. However, by the end of the year, it’s definitely lacking in unique or distinct features. Of course, this should change in early 2025 — we’ll see if Samsung manages to cook up something new and different, which the competition does not yet have.
