Gemma 3n Model Launch and AI Developer Updates for August 2026
This week’s digest covers Gemma 3n’s multilingual multimodal release, new NVIDIA Google Cloud AI builder resources, Haystack v2.3.0 framework comparison, Google Cloud’s Managed Agents API, and a curated list of commercially usable open LLMs.
This week in AI and dev.
A short, hand-picked digest from the frontier. Model drops, new tools, and research worth shipping.
Model Releases
Gemma 3n Released Multilingual Multimodal Model
Gemma 3n offers native multilingual and multimodal capabilities for text image video and audio processing.
Developer Tools
NVIDIA Google Cloud Launch AI Builder Resources
New learning path JAX on GPUs Dynamo codelab and monthly livestreams help developers optimize AI inference.
Research
Haystack v2.3.0 Framework Comparison Guide
Provides side by side evaluation of Haystack and other LLM frameworks for building retrieval augmented pipelines.
Cloud and Infrastructure
Google Cloud Announces Managed Agents API
Managed Agents API lets developers run custom agents in secure Google hosted environments with Agent Platform integration.
Community Picks
Open LLMs List for Commercial Use
Curated list of permissively licensed LLMs including Apache 2.0 MIT and OpenRAIL M models ready for production.
Learn more about our AI services or see recent work in our portfolio.