---
format: "aidr-story-markdown/v1"
id: "56dc36f6d0013c196cbbfb98ee1ef5016332d10d267d10db2862dfb4bdfc7a45"
canonical_url: "https://aidr.today/56dc36f6?lang=en"
title: "Google's First Open Multimodal Embedding Model Powers Local Foresight App"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-07T15:40:09.000Z"
category: "Data"
topics: ["google","open-source","data-engineering","llm"]
source_urls: ["https://huggingnews.com/ai/update-googles-first-open-multimodal-embedding-model-powers-local-foresi-ca476f11","https://huggingface.co/blog/LiquidAI/open-d1","https://x.com/sundarpichai/status/2107501975671890211","https://x.com/sgl_project/status/2107505313482154489","https://x.com/vllm_project/status/2107535469437444467","https://x.com/ollama/status/2107584722465616123","https://x.com/Google/status/2107505123941376129","https://x.com/testingcatalog/status/2107853628656496657"]
summary: "Google launched the Foresight app for macOS, which uses local AI models including EmbeddingGemma 2 to transcribe audio and build knowledge graphs from Google Drive without cloud connectivity. The 740M parameter model unifies text, code, images, video, and audio in a shared embedding space, compressing 10M vectors from 30GB to 0.4GB of RAM using 1-bit TurboQuant and Matryoshka Representation Learning while retaining 94.5% of baseline retrieval quality. AI framework providers including vLLM, Ollama, SGLang, and llama.cpp integrated the model on its first day of release."
---

# Google's First Open Multimodal Embedding Model Powers Local Foresight App

> [Open the canonical story](<https://aidr.today/56dc36f6?lang=en>)

**Published:** 2026-10-07T15:40:09.000Z
**Category:** Data
**Topics:** google, open\-source, data\-engineering, llm

## Summary

Google launched the Foresight app for macOS, which uses local AI models including EmbeddingGemma 2 to transcribe audio and build knowledge graphs from Google Drive without cloud connectivity\. The 740M parameter model unifies text, code, images, video, and audio in a shared embedding space, compressing 10M vectors from 30GB to 0\.4GB of RAM using 1\-bit TurboQuant and Matryoshka Representation Learning while retaining 94\.5% of baseline retrieval quality\. AI framework providers including vLLM, Ollama, SGLang, and llama\.cpp integrated the model on its first day of release\.

## Sources

- [Story source](<https://huggingnews.com/ai/update-googles-first-open-multimodal-embedding-model-powers-local-foresi-ca476f11>)
- [Story source](<https://huggingface.co/blog/LiquidAI/open-d1>)
- [Story source](<https://x.com/sundarpichai/status/2107501975671890211>)
- [Story source](<https://x.com/sgl_project/status/2107505313482154489>)
- [Story source](<https://x.com/vllm_project/status/2107535469437444467>)
- [Story source](<https://x.com/ollama/status/2107584722465616123>)
- [Supporting source](<https://x.com/Google/status/2107505123941376129>)
- [Supporting source](<https://x.com/testingcatalog/status/2107853628656496657>)