HuggingSnap: On-device Visual Assistant

Category

Description of App

Transform your iPhone into a visual assistant with real-time object description and visual understanding powered by on-device AI.

HuggingSnap brings smart vision AI to your iPhone. Using our smolvlm2 vision model, the app understands what your camera sees in real-time without sending any data to the cloud.

Just point your camera, ask a question or request a description, and HuggingSnap will identify objects, explain scenes, read text, and make sense of what you're looking at. It's helpful when shopping, traveling, studying, or just exploring your surroundings.

Key Features:
- Processes everything on your phone - your data stays private
- Works in real-time with no delay
- No internet needed - works offline anywhere
- Easy on your battery
- Reads and translates text in images
- Describes scenes for better accessibility
- Search using your camera
- Choose what types of objects to recognize

HuggingSnap turns your iPhone into a helpful visual companion that sees the world with you!

Terms of use & privacy policy:
Terms of use: https://huggingface.co/terms-of-service
Privacy policy: https://huggingface.co/privacy

Version

1.2

Free or Paid

Free

Apple Watch Support

No

Device(s) App Was Tested On

iPhone

iOS Version

18.4.1

Accessibility Comments

Accessibility with Voiceover was significantly improved in Version 1.1.

VoiceOver Performance

VoiceOver reads most page elements.

Button Labeling

Most buttons are clearly labeled.

Usability

The app is fully accessible with VoiceOver, but the interface could be easier to navigate and use.

Recommendations

1 people have recommended this app

Most recently recommended by Enes Deniz 3 days 15 hours ago

Options

Comments

By Brad on Tuesday, March 25, 2025 - 06:10

For those that do try it, or want another tool, i'd recommend contacting the dev and seeing where things go.

Their email is: [email protected]

By Brian on Tuesday, March 25, 2025 - 06:10

If the developer can get the accessibility sorted, this could potentially be a great addition to the majority of tools that blind folks use on a daily basis. :-)

By Enes Deniz on Wednesday, April 30, 2025 - 11:23

We should have the ability to;
• Select HuggingSnap from the Share sheet or access previously captured photos/videos within the app to have them described
• Know whether the front or back camera is currently selected
• View the size of the downloaded model, delete it and download other models to strike a balance between performance and quality, taking into account memory and hardware requirements
• Ask follow-up questions
• Take advantage of dictation/voice mode using the system voice, with customizable parameters, and even Siri integration, if applicable
• Have OCR functionality in multiple languages and scan PDFs
• Have HuggingSnap describe the content on the screen without having to take a screenshot and export it to HuggingSnap manually, to supplement VoiceOver’s Screen Recognition
• Teach HuggingSnap faces and objects and have them labeled by name wherever they’re encountered
• Explore images/scenes by touch (i.e., by moving the finger around the screen), and get audio cues in 3-D to get a better sense of the position/distance/depth of each object
• Enter a “system prompt” to be used for every description
A couple of notes:
1. Just copied the whole thing from the text field and I don't feel like adding all the HTML tags to convert it to a list, unless requested.
2. I figured out soon enough that we could already ask follow-up questions. Also, I thought of suggesting that we be able to capture photos or videos in portrait or landscape mode after submitting the review.

By Stephen on Wednesday, April 30, 2025 - 13:48

So I just downloaded the app and the buttons are labeled and the app seems to be accessible from my end.