Deertouch X1 Review: I Tried This AI Photo Frame That Turns Photos Into Interactive Characters
Most digital photo frames do one thing: display your pictures. Some can play videos, and others let family members send photos remotely.
The Deertouch X1 is trying to do something quite different.
It’s marketed as an AI Persona Frame, and after spending time with an engineering sample, I can see why. Instead of simply displaying a photo, you can use the DeepSire app to create a personalized AI character based on someone’s appearance, voice, and personality. You can then bring that character onto the physical frame and interact with it.
I was able to customize an Avatar’s relationship, tone, personality, traits, and background. I also generated an animated video, experimented with voice interactions, and created custom actions using ordinary language prompts. No coding required.
There are some genuinely fun ideas here. The Avatar generation quality was better than I expected, and the ability to make a character respond with specific gestures gives the X1 a lot of room for experimentation.
But it’s not all smooth sailing. Responses took around two to three seconds in my testing. Lip movements didn’t always match the speech, and the built-in speaker sounded noticeably artificial. I also ran into an odd login issue between the phone app and the frame.
One thing to keep in mind: I tested a borrowed engineering sample, not a final retail unit. Some hardware and software details may change before the product ships.
Here’s what I found after putting the Deertouch X1 through its paces.
Table of Contents
Specification
| Feature | My Experience |
|---|---|
| Product type | AI Persona Frame |
| Companion app | DeepSire |
| Character customization | Relationship, Tone, Personality, Traits, Background |
| Avatar creation | Static image and video |
| Video generation time | Approximately 2 minutes |
| Video generation cost | 9,000 Coins |
| Custom action cost | 399 Coins |
| Voice interaction | Supported on my test unit |
| Custom actions | Natural-language prompts |
| Display | Touchscreen |
| Battery | No built-in battery on my engineering sample |
| Review unit | Borrowed engineering sample |
What Is the Deertouch X1?
The Deertouch X1 isn’t really trying to compete with a regular digital photo frame.
Its main idea is to let you create a personalized AI character and give that character a place in your home.
The process starts in the DeepSire app. You upload a person’s photo and voice sample, configure their personality, and generate an Avatar. The resulting character can then be displayed on the X1.
You can also interact with the Avatar and trigger different actions through voice conversations.
That’s a pretty different proposition from a frame that simply cycles through family photos.
The broader DeepSire ecosystem talks about digital presence, memories, and personalized AI experiences. I find that direction interesting, although I wouldn’t assume every feature or promise described across the DeepSire ecosystem is already available on the X1.
For this review, I’m sticking to what I could actually use on the device.
Design and Display: Familiar Looks, Decent Screen
At first glance, the X1 looks much like a regular digital photo frame. There’s no particularly unusual design element that immediately tells you it’s an AI device.
The touchscreen was reasonably responsive during my testing, which made navigating the interface straightforward.
The screen itself looked good when viewed from the front. Images were clear, and I didn’t have any major complaints about the viewing experience head-on.
The story was different from the side. Reflections became quite noticeable at an angle, which could be distracting depending on where you place the frame.
There’s also one hardware limitation worth mentioning.
My engineering sample doesn’t have a built-in battery, so it needs to stay plugged into a power source. That makes placement less flexible than a battery-powered device.
I haven’t confirmed whether the final retail version will have the same power arrangement, so I wouldn’t treat this as a confirmed specification for the shipping model.
Overall, the hardware is fairly ordinary. The interesting stuff happens inside the DeepSire app and through the AI features.
Creating an AI Avatar With the DeepSire App
This was where things started to get interesting.
To customize a character, I opened the DeepSire app and went to:
My → My Photo Frame
There are several options for shaping the character:
| Setting | What You Can Customize |
|---|---|
| Relationship | The relationship between the character and the user |
| Tone | How the character communicates |
| Personality | The character’s overall personality |
| Traits | Individual characteristics |
| Background | The character’s background information |
You can also upload a person’s photo and voice sample to customize the Avatar.
I liked having this level of control. Rather than picking a generic AI character and accepting whatever personality it comes with, you get to define several parts of the character yourself.
It also makes the process feel more personal. You’re not just generating another AI face; you’re trying to create a character with a particular identity.
I haven’t finished testing how consistently each setting changes the character’s actual behavior, so I won’t claim that every option works equally well. That’s something I’d like to explore further.
Static Avatar vs. AI-Generated Video
DeepSire offers both static Avatar and video generation.
I tried the video generation feature using an uploaded portrait and voice sample. The process took roughly two minutes and used 9,000 Coins.
The result was surprisingly good.
The generated person looked realistic, and the facial features stayed fairly consistent with the original photo in the video I tested.
That’s worth pointing out because a realistic animation isn’t much use if the person suddenly looks like someone else halfway through.
My initial result was encouraging, although I’d want to test more photos and repeat the same generation several times before making a broader claim about consistency.
For now, I’d say the Avatar generation feature is one of the stronger parts of the X1 experience.
The Fun Part: Making Your AI Avatar Move and Respond
This is probably my favorite feature of the Deertouch X1 so far.
The frame doesn’t just display an Avatar or play a pre-generated video. You can also interact with the character and trigger specific actions through conversation.
There are several preset actions available in the interface, including:
Thumbs Up
Gentle Fingertip Touch
Calmly Crossing the Arms
Raise a Hand in Greeting
Nod inside the bag
Genlt Smile While Getting a Haircut
I’ve kept the last two names as they appeared in my test notes. They may contain wording or spelling issues in the engineering software.
The actions give the Avatar a little more personality. Instead of just responding with words, it can also react with gestures.
Creating Your Own Actions Without Coding
The custom action feature is even more interesting.
DeepSire provides two types of prompts:
Call Wake Action Prompts
These define the trigger words or phrases that activate an action.
Action Prompts
These describe what the Avatar should actually do.
Both can be written in natural language. You don’t need to know how to code or understand any particular programming language.
For example, you could try instructions like:
When I say hello, smile and nod.
When I say good morning, raise your hand and wave.
When I say let’s think about it, calmly cross your arms and smile.
These are examples of the kind of instructions you could experiment with, rather than verified results for every prompt.
Creating a custom action cost 399 Coins on my test unit.
I enjoyed playing around with this feature more than I expected. It gives you a way to make the character feel a little more like your own creation, rather than just another animated face on a screen.
There’s also something appealing about being able to describe an action in plain English and see it happen without having to deal with code.
I’d still like to test how reliably custom actions trigger and whether the same action behaves consistently across repeated attempts.
But as a concept, this is one of the features that sets the Deertouch X1 apart from a standard AI photo frame.
Talking to the Avatar: Fun, But Not Quite Natural Yet
During my testing, I could interact with the Avatar through voice, and the frame responded with speech and physical movements.
The movements themselves looked reasonably natural. That was a pleasant surprise.
However, I noticed a delay of around two to three seconds after finishing a spoken interaction before the frame responded.
That’s not an enormous wait, but it’s noticeable when you’re looking directly at a character and expecting it to react.
A phone-based chatbot can get away with a little waiting time. A character sitting on a desk feels different. You naturally expect it to respond more like something you’re having a conversation with.
Lip Sync Needs Work
Another issue was lip synchronization.
The Avatar’s mouth movements didn’t always line up with the speech. When you’re watching a personalized character talk, even small timing differences can make the experience feel less convincing.
I’d like to see this improved in the final software.
The Built-in Speaker Is Another Weak Spot
The speaker quality wasn’t particularly impressive either.
The sound had a noticeable synthetic, mechanical quality. It didn’t sound as natural as I’d hoped, especially considering how much emphasis the product puts on personalized voices.
This is an area where better hardware or audio tuning could make a real difference.
After all, creating a realistic-looking Avatar is only half the job. If the voice sounds artificial and the mouth movements don’t match, the overall illusion falls apart pretty quickly.
Here’s how I’d sum up the interaction experience so far:
| Area | My Experience |
|---|---|
| Response time | Around 2–3 seconds |
| Physical movements | Reasonably natural |
| Lip synchronization | Needs improvement |
| Speaker quality | Noticeably mechanical |
| Custom actions | Fun and flexible |
How Much Does the DeepSire AI Generation System Cost?
The AI features are fun to experiment with, but there’s another part of the experience worth looking at: the cost of generating content.
DeepSire uses a coin-based payment system.
New User Free Generation
After registering a new account, I found that new users receive one free AI persona generation opportunity.
This can be used to try both static Avatar and video Avatar generation.
I didn’t find additional free Coins or other new-user benefits during my testing.
It’s a nice way to let people try the main feature before spending money, although it’s worth remembering that this is a one-time opportunity rather than unlimited free generation.
DeepSire Coin Packages
Here are the recharge options I found in the app:
| Price | Coins |
|---|---|
| $6.99 | 3,480 |
| $9.99 | 4,980 |
| $28.99 | 14,580 |
| $96.99 | 48,880 |
| $137.99 | 69,880 |
| $219.99 | 111,880 |
The larger packages offer slightly better value per Coin, but the difference isn’t huge.
What Does One AI Avatar Video Actually Cost?
In my test, generating one Avatar video took about two minutes and consumed 9,000 Coins.
Based on the available packages, those 9,000 Coins represent roughly $17.70–$18.08 in proportional value.
However, that’s not necessarily what you’ll pay at checkout.
If you start with zero Coins, the smallest package that covers one 9,000-Coin generation costs $28.99. You’ll receive 14,580 Coins, leaving 5,580 after generating one video.
That’s an important distinction when budgeting for the experience.
What About Custom Actions?
Creating a custom action costs 399 Coins on my test unit.
That’s approximately $0.78–$0.80 in proportional value based on the available packages.
The smallest recharge package is $6.99, so users without an existing balance will need to spend more upfront.
Do You Pay Again When Changing an Avatar?
Here’s something I liked about the system.
I could modify an existing Avatar’s personality settings and voice without generating the Avatar again or paying another generation fee.
That’s a welcome detail. It means you can experiment with certain aspects of an existing character without having to pay for a completely new generation every time.
However, generating a new image or video is a separate process and may consume Coins.
I haven’t confirmed the exact cost of static Avatar generation or whether failed generations are refunded.
DeepSire Pricing Summary
| Feature | Coins | Status |
|---|---|---|
| Initial free generation | One free opportunity | Tested |
| Static Avatar generation | Not confirmed | Further testing needed |
| Video Avatar generation | 9,000 | Tested once |
| Custom action | 399 | Tested |
| Personality changes | No additional generation charge | Tested |
| Voice changes | No additional generation charge | Tested |
| Refund for failed generation | Unknown | Not tested |
My advice is to look beyond the initial free generation when considering the long-term cost.
If you only want to create one character and occasionally experiment with it, the free opportunity gives you a way to get started.
If you enjoy creating multiple Avatars and generating videos, the Coins system becomes a more important part of the ownership cost.
The App and Frame Don't Always Play Nicely Together
I ran into an unexpected issue when moving from the DeepSire app to the physical frame.
After creating an Avatar in the mobile app, I switched to the X1 and logged in using the same account.
I could see the generated Avatar on the frame, so the basic content transfer worked.
However, the moment I logged into the frame, the mobile app automatically logged out and returned to its login screen.
I tried both password-based login and QR-code login on the frame, and this was the behavior I encountered.
That was confusing.
The app is where I create and customize my Avatars, while the physical frame is where I display and interact with them. I’d naturally expect to use both devices without having to keep switching accounts.
I haven’t confirmed whether this is an intentional single-session restriction or an issue with the engineering sample.
It’s something I’d like Deertouch to clarify.
Ideally, both devices should stay logged in, with changes made in the app syncing smoothly to the frame.
This might sound like a small software issue, but it can make a noticeable difference in everyday use.
What I Like About the Deertouch X1 (And What Needs Work)
After spending time with the engineering sample, I have a fairly clear picture of where the product stands.
We Like
The Avatar customization is genuinely interesting.
Being able to define a character’s relationship, tone, personality, traits, and background gives users more control than simply choosing a preset AI character.
The generated video looked good in my test.
The result was reasonably realistic, and the person’s facial identity remained consistent.
Custom actions add a lot of fun.
I particularly like being able to define trigger phrases and actions using natural language. There’s no need to learn programming just to experiment with different character behaviors.
Changing an existing character doesn’t require another generation.
At least for the personality and voice adjustments I tested, I didn’t have to pay for another generation.
We Don't Like
The interaction could be faster.
A two-to-three-second response delay isn’t terrible, but it makes conversations feel less natural.
Lip synchronization needs improvement.
The mouth movements don’t always match the speech.
The speaker sounds too mechanical.
A more natural-sounding speaker would help the Avatar feel more convincing.
The app and frame need better integration.
Having the mobile app log out when the frame signs in isn’t a particularly convenient experience.
The screen could handle reflections better.
The display looks good from the front, but reflections are more noticeable from the side.
The engineering sample needs a power connection.
My unit doesn’t have a built-in battery. I’ll need to see whether the retail version is different.
Is the Deertouch X1 Worth Buying?
I wouldn’t judge the Deertouch X1 by the same standards I’d use for a regular digital photo frame.
If all you want is a screen to display family photos, there are simpler options that don’t require an AI credit system or character setup.
The X1 is aimed at people who want to experiment with something different: personalized AI characters that can appear on a physical screen, speak, and respond with customized actions.
I can see the appeal, particularly for AI enthusiasts and people who enjoy experimenting with new technology.
The custom action feature is probably the most unexpected part of my experience. Being able to describe a gesture in plain language and connect it to a trigger phrase makes the device much more entertaining than I initially expected.
But I also think the software and hardware need more work.
The response delay, speaker quality, lip synchronization, and account-login behavior all affect the experience. They’re not necessarily deal-breakers for an engineering sample, but they’re things I’d want addressed before recommending the final retail version without reservations.
My current verdict: The Deertouch X1 is a promising AI Persona Frame with plenty of room for creativity. The character creation and custom action features are its strongest selling points. The interaction still needs polishing, though, and I’d like to see how much the final production version improves before making a firm buying recommendation.
The question I keep coming back to is this: after the excitement of creating your first AI character wears off, will you actually want to keep talking to it?
That’s what will ultimately determine whether the Deertouch X1 is more than an interesting AI gadget.
Frequently Asked Questions
A regular digital photo frame displays pictures and videos. The Deertouch X1 adds personalized AI Avatars, voice interaction, personality settings, and custom character actions.
The X1 uses the DeepSire app for creating and customizing AI Avatars. The path I used was My → My Photo Frame.
Yes. The DeepSire app lets you upload a person’s photo and voice sample and configure settings such as Relationship, Tone, Personality, Traits, and Background.
Yes. My test unit supported both static Avatar images and generated Avatar videos.
My test took approximately two minutes. The generation consumed 9,000 Coins.
Yes. The system provides Call Wake Action Prompts and Action Prompts. Both can be written in natural language, so you don’t need programming experience to describe an action and its trigger.
Creating a custom action cost 399 Coins on my engineering sample.
New users receive one free AI persona generation opportunity. I was able to try both static and video Avatar generation through the new-user benefit.
My engineering sample did not have an internal battery and needed to remain connected to power. I haven’t confirmed whether the final retail version will be the same.
If you’re interested in personalized AI characters and interactive hardware, it’s an interesting product to watch. However, I’d like to see improvements to response time, audio, lip synchronization, and account management before making a final recommendation.
Review Disclosure
I tested a borrowed Deertouch X1 engineering sample. I did not purchase the review unit.
The opinions and observations in this article are based on my own hands-on testing. The final retail hardware, software, and features may differ from the engineering sample.
If this article contains affiliate links, I may earn a commission from qualifying purchases at no additional cost to you.