What's Changed

• The app is now accessible.
• Migrated to the LiteRT-LM Kotlin API. Check out the doc here.
• Added Qwen2.5-1.5B-Instruct, Phi-4-mini-instruct, and DeepSeek-R1-Distill-Qwen-1.5B litertlm models for you to try in AI Chat and PromptLab.
• Bug fixes about broken scrolling on long messages and partial message copying, and other minor fixes and improvements.

Full Changelog: 1.0.7...1.0.8

What's Changed

• Fix NPE crash due to failure to download. by @copybara-service[bot] in #282
• Remove object instead of index to avoid potential crash. by @copybara-service[bot] in #283
• Update splash screen with animated icon. Also make the "failed to load allowlist" dialog dismissable. by @copybara-service[bot] in #287
• [gallery] fix a crash on user providing invalid text value for a number by @copybara-service[bot] in #294
• Add animation for audio recording and simplify audio recording ui. by @copybara-service[bot] in #296
• Hide keyboard after sending a prompt; Scroll to bottom when a new prompt is sent; Show warning when user tries to import a web model. by @copybara-service[bot] in #301
• [gallery] allow download model using a git reference by @copybara-service[bot] in #310
• Fixes a typo by @EsmaeelNabil in #261
• Moro auto-scrolling related fix. by @copybara-service[bot] in #313

New Contributors

• @EsmaeelNabil made their first contribution in #261

Note on accessibility: We've heard your valuable feedback regarding the lack of accessibility in the app, and we are actively working on making the app fully accessible by our next release. Thank you so much for your continued support and patience!

Full Changelog: 1.0.6...1.0.7

Minor code clean up and refactoring

Refactored the model download logic. This should fix various model download issues reported from the community.

The beta version of the Google AI Edge Gallery app is now live on Google Play! It's currently available for download in the US only.

• For US users: You can get the app from Google Play.
• For users outside the US: Please download the attached APK to install the app. Please stay tuned for the global launch soon.

Release notes:

• Added multi-image multi-turn support: In Ask Image, you can now upload multiple images (up to 10) and ask questions about the images in multiple turns.
• Added audio support in Audio Scribe: Upload or record an audio clip, and ask the model to transcribe it into text or translate it into another language.
• UI polish.
• Fixed various crashes and usability issues reported by the community.

Fix some potential memory leak in image capture sheet

• Use camerax to implement image capture in-app instead of using the native camera app to reduce the possibility of app being killed due to GPU OOM.
• Show accelerator name in chat message sender labels.
• Attach download workers to silent foreground notifications to make them less likely to be killed.
• Update app icon to be consistent with Google style.

• Update Gemma 3n 4B model.
• Update context length to 4K for Gemma 3n models.
• Fix a bug related to LLM inference engine cleanup.
• Various UX improvements.

• Fix a download resume bug