18 stories tagged gemma, newest first.
103 points by armcat | 19 comments
BottleCap AI blog post announcing ThinkingCap-Gemma-4-26B-A4B, a fine-tune cutting thinking tokens by 39% on average with near-unchanged benchmark accuracy.
395 points by ilreb | 42 comments
Google blog post announcing EmbeddingGemma 2, an open lightweight on-device model that embeds text, images, audio, and video into a unified space.
538 points by firelex | 205 comments
GitHub repository for Jeff, a project that fine-tunes Qwen3.5 and Gemma 4 small models for fast zero-shot classification decisions.
72 points by jamie-simon | 9 comments
Blog post by Jamie Simon on high-dimensional statistics, discussing a paper by Dhruva Karkada on data statistics and periodic structure in embeddings like calendar months.
130 points by allanrbo | 40 comments
Blog post describing a single-function Jev-like wrapper for querying LLMs and vision models, demonstrated with a webcam script scoring frames via Gemma.
845 points by gitpusher42 | 300 comments
GitHub repo for TurboFieldfare, a Swift/Metal inference engine running 4-bit Gemma 4 26B-A4B on M-series Macs in about 2 GB RAM by streaming experts from SSD.
1490 points by jfb | 569 comments
Vicki Boykis blog post arguing that running LLMs locally on a 64GB M2 Mac is now practical, describing models like Mistral 7B, Gemma 3, OpenAI OSS-20B and Qwen 3.
464 points by kkm | 114 comments
A blog tutorial on running Gemma 4 and Qwen3.6 locally on macOS with llama.cpp, MTP speculative decoding, and multimodal support as a coding agent.
1007 points by rvz | 379 comments
Google blog post announcing Gemma 4 12B, an encoder-free multimodal model designed to run on laptops with mobile-first efficiency and advanced reasoning.
451 points by asenna | 127 comments
Blog post describing how the author used Gemma 4 running locally on a five-year-old MacBook to index a year of personal video footage overnight.
829 points by janandonly | 226 comments
Dutch App Store listing for Google AI Edge Gallery, a free iPhone app for running on-device LLMs like Gemma 4 for chat, image questions, and agent skills.
1705 points by jeffmcjunkin | 450 comments
Google DeepMind product page announcing Gemma 4, a family of open models built from Gemini 3 research, with links to docs and Google AI Studio.
475 points by meetpateltech | 193 comments
Google Developers Blog post announcing Gemma 3 270M, a compact open model designed for hyper-efficient, task-specific AI.
361 points by bundie | 159 comments
Google for Developers blog post introducing Gemma 3n, a developer guide to Google's new on-device open model.
418 points by meetpateltech | 145 comments
Google for Developers blog post announcing the preview of Gemma 3n, an efficient, mobile-first open AI model designed for on-device use.
525 points by prettyblocks | 271 comments
Ask HN thread asking what people have built with tiny 0.5B–3B parameter language models, such as those available on Ollama.
387 points by mfiguiere | 127 comments
GitHub repository for gemma.cpp, Google's lightweight standalone C++ inference engine for running Gemma language models.
1075 points by meetpateltech | 476 comments
Google blog post announcing Gemma, a family of open AI models built from the same research and technology as Gemini.