Type: url
Original: https://developers.googleblog.com/introducing-support-for-local-ai-models-in-the-antigravity-sdk/
Status: generated
Summary
Google announces that the Antigravity SDK supports agentic workflows on local models, with initial support for Gemma 4 26B A4B running on Google AI Edge’s LiteRT and plug-in support for OpenAI-compatible servers such as Ollama, LM Studio and vLLM. The post gives cost, privacy, offline resiliency and hybrid cloud-local workflows as the reasons to run agents locally, and shows a hybrid Architect-Builder demo in which a cloud Gemini model plans a security-patching task from filenames alone while local Gemma instances audit, fix and validate the code on-device.