#LiteRT
LightNews LightNews
iben.ch
Mondjuk azt nem gondoltam teljesen at, h ez a dog olyan 30 litert fogyaszt szazon. Orditott a V8 a Big Sur szerepentineken. 174 litert tankoltam, amikor negyedig volt. 😀
October 7, 2026 at 1:38 PM
pub.dev.web.brid.gy
v1.11.2 of flutter_edge_ai_litertlm
LiteRT-LM (.litertlm) engine for flutter_edge_ai: dart:ffi on 5 native platforms plus web, CPU/GPU/NPU, and the LiteRT embedding backend. Changelog excerpt: - Allow `code_assets`2.x, which other native plugins now require (#603).
pub.dev
October 10, 2026 at 3:46 PM
babynair.bsky.social
ML Drift: Google’s Next-Gen GPU Engine for On-Device AI Inference
Learn how Google ML Drift accelerates on-device AI inference with GPU optimization, LiteRT, cross-platform support, and next-generation edge AI.

seoinsights.co.in/ml-drift-gpu...
October 9, 2026 at 2:15 PM
ruen.bsky.social
You can do that and decently well on purely CPU.

Google showed off an offline translator powered by a somewhat tuned version of Gemma E2B running on a Pi 5.

www.raspberrypi.com/news/masteri...
Mastering edge AI on Raspberry Pi with LiteRT and Gemma - Raspberry Pi
Our friends at Google have been busy working on improving LiteRT performance on the Raspberry Pi platform.
www.raspberrypi.com
October 11, 2026 at 1:18 AM
breachprotocol.bsky.social
Google’s Apache-2.0 ML Drift engine supports four graphics interfaces for on-device inference and can run standalone or accelerate LiteRT.
Google releases ML Drift for graphics-processor inference on devices
Google’s Apache-2.0 ML Drift engine supports four graphics interfaces for on-device inference and can run standalone or accelerate LiteRT.
groundtruth.day
October 10, 2026 at 6:37 AM
tech-trending.bsky.social
Keras入門　第４回　Kerasモデルの量子化(LiteRT)
https://zenn.dev/hoshinagi1219/articles/6c2446701595cc
Keras入門　第４回　Kerasモデルの量子化(LiteRT)
zenn.dev
October 7, 2026 at 12:22 AM
dailyzenntrends.bsky.social
今日のZennトレンド

Keras入門　第４回　Kerasモデルの量子化(LiteRT)
Google Colaboratory環境でai-edge-litertライブラリを使用し、Kerasで作成したMNISTモデルを読み込んでLiteRTで動作させる方法を解説した記事です。
Kerasе…Ҙй–ҖгҖҖ第４еӣһгҖҖKerasгғўгғҮгғ«гҒ®йҮҸеӯҗеҢ–(LiteRT)
星杜なぎさです。今回は、Kerasモデルの量子化についてまとめてみました。 環境google colaboratoryPython : 3.13.15keras：3.13.2tensorflow：2.21.0ai-edge-litert：2.2.0 1. 準備 1.1 ライブラリの読み込み2026年10月6日現在、google colaboratoryにはai-edge-litertのライブラリが
zenn.dev
October 7, 2026 at 12:33 AM
webai.bsky.social
So @Google's #EmbeddingGemma2 model just dropped - naturally we have a #WebAI port on client side in JS - totes private, zero cost other than your own power, and low latency right in browser! Try the demo - Dare you to use mic & meow and see what you get! google-ai-edge.github.io/LiteRT-LM/we...
October 7, 2026 at 5:35 AM
fabricatedtechnobabble.com
I forgot to share my This Week in Open Source post from last week on the @opensource.google blog.

opensource.googleblog.com/2026/10/this...
This Week in Open Source for October 2, 2026
Explore Gemma 4 12B inference math, local RAG with LiteRT, Google's EnvHarness, Mozilla & Mila's open AI stack, and October OSS events.
opensource.googleblog.com
October 5, 2026 at 12:49 PM
The latest edition of This Week in Open Source (#18) from @opensource.google covers Gemma 4 12B inference, 100% local RAG (LiteRT + Qdrant), adaptive agent training with EnvHarness, @mozilla.org & Mila's open AI stack, and Oct/Nov events: https://goo.gle/twios-18 #OpenSource
This Week in Open Source for October 2, 2026
google.smh.re
October 6, 2026 at 7:21 PM
pub.dev.web.brid.gy
v0.5.4 of flutter_edge_ai_speech
On-device speech (STT + TTS) for flutter_edge_ai via the LiteRT C API + dart:ffi. Opt-in Stt/TtsBackendProvider. Changelog excerpt: - Require `flutter_edge_ai`^2.1.0; README covers TTS from assets, files and bundles.
pub.dev
October 6, 2026 at 9:43 PM
opensource.google
This Week in Open Source #18 is live! 🌐✨ 🧮 Inside Gemma 4 12B inference & 100% local RAG (LiteRT + Qdrant) 🧪 Adaptive agent training w/ EnvHarness 🏗️ @mozilla.org & Mila's open AI stack ⚠️ The maintainer cost of LLM "vibe-coding" 🗓️ Oct/Nov events: goo.gle/twios-18 #OpenSource
Google Open Source banner: This Week in Open Source #18 (Oct 2, 2026). Headline: What happens when open source AI moves from benchmarks to infrastructure? Featuring Gemma 4 local RAG, EnvHarness agent sandboxes, and Mozilla & Mila's open stack. Card 2 of 4: Open Source AI & Infrastructure highlights. Features Gemma 4 12B KV-cache sizing math and offline LiteRT+Qdrant RAG, Google's EnvHarness adaptive RL agent sandboxes, and Mozilla & Mila Quebec's sovereign open AI ecosystem. Card 3 of 4: Community & Sovereignty highlights. Features maintainer costs of unreviewed AI "vibe-coding" PRs, UK Government open-source digital sovereignty stewardship, and native FreeBSD-to-Linux syscall translation on PlayStation 5. Card 4 of 4 (Bluesky): Call to action pointing to goo.gle/twios-18 in post above. Highlights reads from @mozilla.org, Mila, @qdrant.to, LiteRT, and Gemma 4, plus Oct/Nov events including ValkeyConf, MCP Dev Summit, and OSS Europe.
October 2, 2026 at 6:35 PM
geshan.com.np
Introducing Support for Local AI Models in the Antigravity SDK
Introducing Support for Local AI Models in the Antigravity SDK
Build privacy-first AI agents with the Google Antigravity SDK. Now supporting offline, local workflows via LiteRT, Gemma 4 26B, and hybrid orchestration.
developers.googleblog.com
October 2, 2026 at 8:30 AM
tom-doerr.bsky.social
Runs a fully offline voice translator on a Raspberry Pi using Gemma 4 and LiteRT-LM.
Screenshot of the repository
September 28, 2026 at 5:34 PM
premarcu.bsky.social
Tool calling is the real test. A malformed call on a 26B local costs a full re-prefill unless prefix cache holds, so I constrain decoding instead of retrying. Does LiteRT-LM expose grammar hooks, or is that still a proxy I write?
September 27, 2026 at 5:29 PM
opensource.google
When creators of a proprietary agent tool build first-class support for open-weights Gemma 4 26B, LiteRT-LM, vLLM & Ollama, the signal is clear: real-world AI needs local & open source in the mix. w/ @developers.google.com @ntaylormullen.bsky.social @rodydavis.com https://goo.gle/agy-local-ai
September 25, 2026 at 5:02 PM
opensource.google
Gallery links (by Sachin Kotwani & @ntaylormullen.bsky.social):
• Main article: https://goo.gle/agy-local-ai
• Hybrid Gauntlet: https://goo.gle/47cKYyV
• Python SDK: https://goo.gle/agy-sdk-local
• Gemma 4 26B: https://goo.gle/gemma4-26b-litert
• LiteRT-LM: https://goo.gle/litert-lm-docs
September 25, 2026 at 5:02 PM
alessiopomaro.it
🧠 #Google ha annunciato che l'#Antigravity SDK ora supporta i workflow locali: le stesse capacità agentiche, ma eseguite interamente sulla propria macchina, con supporto iniziale per #Gemma 4 26B A4B tramite LiteRT di Google AI Edge.
👉 I dettagli: lnkd.in/p/ehAySBUW

#AI #GenAI #GenerativeAI #LLM
September 26, 2026 at 10:48 AM
mammutreal.bsky.social
nálunk csak a lidl-ben meg egy borboltban kapni a nagy lajoson, de az utóbbi 3400-ért adja, több mint ezerrel drágábban, és a múltkor 1,5 napot sem bírt, pedig ugyanúgy napi 5-6 litert csináltunk csak, mint a lidl-ssel, ami 1 hétig, másfél hétig bírja, szóval oda többé nem megyek 🤷
September 26, 2026 at 1:47 PM
ivantokar.bsky.social
Google’s Antigravity SDK now supports offline local agents. The interesting part is hybrid routing: cloud planning while source code stays with local models.

https://developers.googleblog.com/introducing-support-for-local-ai-models-in-the-antigravity-sdk/

#AI #AIAgents
Introducing Support for Local AI Models in the Antigravity SDK- Google Developers Blog
Build privacy-first AI agents with the Google Antigravity SDK. Now supporting offline, local workflows via LiteRT, Gemma 4 26B, and hybrid orchestration.
developers.googleblog.com
September 27, 2026 at 12:30 PM
babynair.bsky.social
Google Antigravity SDK: Run Local AI Models Offline with Gemma 4
Learn how the Antigravity SDK supports local AI models, Gemma 4 26B and LiteRT for offline agents, privacy-focused workflows and hybrid AI development.
seoinsights.co.in/google-antig...
Google Antigravity SDK
September 24, 2026 at 2:56 PM
mechahara-san.bsky.social
GoogleのAntigravity SDKにローカルモデル対応が来て、Gemma 4（26B）使って完全オフラインで自律エージェント動かせるようになったの熱い！LiteRT経由だからAPIキーもネットも不要で、手元のPC内でツール実行やコード編集が完結するのロマンある〜！
September 25, 2026 at 9:03 PM
k-i.bsky.social
“ Antigravity SDKを使用して、Gemma 4などのオープンモデルを完全にオフラインで実行できます。

Google AI EdgeのLiteRTを基盤としており、Gemma 4をローカルのGPU上で直接動作させることが可能です。APIコストはゼロ、データのプライバシーは完全に保護され、インターネット接続も不要です。 ”

x.com/antigravity/...
Google Antigravity (@antigravity) on X
Run open models like Gemma 4 completely offline in the Antigravity SDK. Built on Google AI Edge’s LiteRT, you can now run Gemma 4 directly on your local GPU. Zero API costs, total data privacy, and no internet required.
x.com
September 24, 2026 at 7:42 AM
glaforge.dev
The #antigravity team announced running local Gemma models on-device with LiteRT.

Good news for #Java developers: you can do this in Java too!

Antigravity Java SDK v0.2.17 is out, reaching full parity with Python SDK v0.1.18.

Walkthrough 🧵👇
glaforge.dev/posts/2026/...
x.com/antigravity...
Google Antigravity (@antigravity) on X
Run open models like Gemma 4 completely offline in the Antigravity SDK. Built on Google AI Edge’s LiteRT, you can now run Gemma 4 directly on your local GPU. Zero API costs, total data privacy, a…
x.com
September 23, 2026 at 10:15 PM
glaforge.dev
Step 1: LiteRT-LM setup

Install the on-device runtime:
pip install litert-lm

Download a Gemma model:
litert-lm import --model-id litert-community/gemma-4-E2B-it-litert-lm --output-dir ~/.litert-lm/models/gemma4-e2b/

Serve locally:
litert-lm serve --port 9379
September 23, 2026 at 10:15 PM