Google Unveils Coral Board for Local AI and Edge Computing
…A major highlight of the launch was Google's demonstration of the Gemma 3 270M language model running locally on the Coral Board. By performing inference on-device, the system can continue…
Search
All sources
xda-developers.com
52
blogs.nvidia.com
13
androidauthority.com
9
huggingface.co
8
cloud.google.com
7
developer.nvidia.com
6
android-developers.googleblog.com
5
theregister.com
4
9to5google.com
4
arstechnica.com
3
extremetech.com
3
wccftech.com
3
Tracked topic
Google Gemma
Gemma is a family of open-weight language models released by Google for text generation and related NLP tasks.
Videos
More videos
Gemma 4 12B Demo: Native Audio Processing in Google AI Edge Eloquent
Bring the power of on-device AI to life with Google AI Edge and Gemma
Gemma 4 Hits 200M+ Downloads: 3 Amazing Local Builds
Building android apps with Gemma 4 for AI coding assistance
Google just casually disrupted the open-source AI narrative…
DeepMind’s New AI: A Gift To Humanity
How to build on-device AI with Gemma 4
Build intelligent Android apps with Google's AI
Google Cloud Live: Accelerate data science and analytics with GPUs
Top 3 AI on Android updates for building intelligent experiences (Google I/O 2026)
Top stories
I tested Gemma 4, Qwen 3.5, and Ministral 3 for vision tasks, and only one understood the assignment
4d agoHugging Face and Cerebras bring Gemma 4 to real-time voice AI
1w ago
aws.amazon.com Amazon SageMaker AI now supports serverless model customization for Gemma 4 models - AWS
1w agoGemma 4's smallest model runs on 3GB of VRAM, and it's the one I actually reach for
1w agoDiffusionGemma is Google's fastest AI yet, but it comes with a big trade-off
…Output quality is still inferior to Gemma 4, so it’s more of an experimental tool than a finished product. Google has released DiffusionGemma, an experimental AI model that takes a very…
NVIDIA Delivers Day-1 Support For DeepMind's DiffusionGemma Open Model Across RTX & DGX Platforms, 150 Tokens/s With DGX Spark
…Gemma 4: DiffusionGemma is built on Gemma 4, a 26-billion-parameter mixture-of-experts model that activates just 3.8 billion parameters per step, pairing a diffusion head with Google’s…
Gemma Scope: helping the safety community shed light on the inner workings of language models
["google-gemma"]
Discussions and forums
New Google Gemma 4 12B Claims Near-26B Performance - We Tested Both!
We ran both models locally on one RTX 4090 and gave each the same task: write a self-contained HTML5 canvas animation with real physics in one file without libraries. Three scenes - a Galton board, two blocks colliding o…
Google's new Gemma 4 12B model is designed to run on any laptop with 16GB of RAM | Gemma 4 12B uses a new encoding scheme and token prediction to punch above its weight.
Google's new Gemma 4 12B model is designed to run on any laptop with 16GB of RAM | Gemma 4 12B uses a new encoding scheme and token prediction to punch above its weight.
Google's new Gemma 4 12B model is designed to run on any laptop with 16GB of RAM
Google's new Gemma 4 12B model is designed to run on any laptop with 16GB of RAM
12
Gemma 4 MTP released
Blog post: https://blog.google/innovation-and-ai/technology/developers-tools/multi-token-prediction-gemma-4/ MTP draft models: https://huggingface.co/google/gemma-4-31B-it-assistant https://huggingface.co/google/gemma-4-…
Gemma 4 with quantization-aware training
Google's collections: https://huggingface.co/collections/google/gemma-4-qat-q4-0 https://huggingface.co/collections/google/gemma-4-qat-mobile And Unsloth's: https://huggingface.co/collections/unsloth/gemma-4-qat Unsloth'…
I tested my local LLM across 4 different tools, and only one actually unleashed its potential
["google-gemma","lm-studio"]
I dumped 650K DNS records into Gemma, and it found patterns my dashboard completely missed
["google-gemma","adguard-home"]
NVIDIA Accelerates Google DeepMind’s DiffusionGemma for Local AI
…Gemma 4: DiffusionGemma is built on Gemma 4, a 26-billion-parameter mixture-of-experts model that activates just 3.8 billion parameters per step, pairing a diffusion head with Google’s…
I tested 3 local LLMs on my RTX 4070 Ti for real work — only one earned a permanent spot
…Google's on-device optimization showing up at runtime. I noticed the difference on the first prompt itself. Where DeekSeek was giving me vague guidance and Qwen an unnecessary essay, Gemma actually…
I stopped running the biggest local LLM that could fit, and a 2B model handles 90% of what I need
…Meet the small models that punch above their weight They're built for regular hardware Gemma 4 E2B is part of Google's Gemma 4 family that dropped at the end of…
I tried Google's new DiffusionGemma, and watching it generate text like an image is unlike any local LLM
…My Mac slowed down system-wide while it was running, and it didn't feel faster than running Google's regular Gemma 4 26B-A4B locally. Google warns that Apple Silicon Macs…