Google DeepMind has released Gemma 4, a family of open-weight generative AI models that includes edge-optimized variants capable of running entirely on a phone or laptop with no internet connection, no API keys, and no server costs after download. The launch, made available under the permissive Apache 2.0 license, pushes lightweight local AI beyond simple chatbots into agentic territory, adding multi-step planning, tool calling, and multimodal processing that all execute on the user's own hardware.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.