Google's Antigravity SDK now runs Gemma 4 agents fully offline with zero token cost
leslysandra · x · 2026-09-25
Google announced that its Antigravity SDK now supports local execution of agentic workflows using Gemma 4 and LiteRT, letting developers run agents entirely on their own machines.
Key points:
- Zero token cost: inference happens fully on-device
- Data privacy: nothing leaves the local machine
- Offline reliability: agents work without connectivity
The SDK also supports OpenAI-compatible endpoints, so Gemma can be served via Ollama, llama.cpp, vLLM and more. Install with pip install google-antigravity litert-lm.
More from coding & agent
- Building a dynamic model router with LangChain and Jev middleware — Arindam_1729 · 2026-09-25
- Agentic RL inside the harness: how LFM2.5 trains with sandboxes and tools — helloiamleonie · 2026-09-25
- On-device agents need 3 things: ~4GB usable RAM, prefill speed, consistent quality — helloiamleonie · 2026-09-25
- Codex user reports quota bug: scheduled tasks burn 30% of $200 plan overnight — sethlazar · 2026-09-25
- The one-shot SaaS experiment seems to be working, reports indie dev gregmushen — gregmushen · 2026-09-25
- Claude Code users urged to run '/claude-api prompt-audit' to tune configs for Opus 5.5 — JeremyNguyenPhD · 2026-09-25