● Covered by 1 source · 1 reportMedium impact1 neutral
Google launches EmbeddingGemma 2 model and AI Edge Foresight Mac app for offline notetaking
Key points
- EmbeddingGemma 2 is a 740M parameter model based on Gemma 4.
- It runs locally, requiring ~191MB RAM for text-only on Pixel 11 Pro.
- AI Edge Foresight Mac app uses the model for offline notetaking.
- The app processes audio and transcripts on-device, not in the cloud.
EmbeddingGemma 2 Model Release
Google announced the release of EmbeddingGemma 2, an AI model designed for on-device processing. This 740 million parameter model is built on the Gemma 4 architecture and is available under an Apache 2.0 license. It is intended to operate within tight resource constraints on consumer hardware.
On-Device Performance
EmbeddingGemma 2 is optimized for local execution. When quantized, the text-only weights require approximately 191MB of active RAM on a Google Pixel 11 Pro, while the full multimodal model uses about 567MB. This allows for tasks like searching through audio recordings based on text queries to be performed directly on the device.
Introducing AI Edge Foresight for Mac
To demonstrate the capabilities of EmbeddingGemma 2, Google launched AI Edge Foresight, a new Mac application. This notetaking app operates by listening in the background during meetings and converting shorthand notes into polished notes using the meeting transcript. All processing of transcripts and audio occurs entirely on the device, without cloud involvement.
Enhanced Information Retrieval
AI Edge Foresight also offers features for information retrieval. Users can direct the app to project folders, reference materials, diagrams, and calendars, including both local files and Google Drive content. The app can then answer questions and provide information based on this data, automatically understanding the content without manual organization. It also includes a conversational retrieval and summary chat feature and Live Assistance for automatic question answering.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
Reporting from
Google released EmbeddingGemma 2, a 740 million parameter AI model, designed for on-device information processing. The company also launched AI Edge Foresight, a Mac application that uses EmbeddingGemma 2 to convert shorthand notes into polished notes from offline recordings and to answer questions based on local files.