Google launches FunctionGemma 270M edge AI model

PANews, December 22 – According to the official Google blog, Google has released FunctionGemma, which specifically fine-tunes Gemma 3 270M as a function-calling model for local/offline agent scenarios. Its features include unified chat and tool execution, support for customized fine-tuning (Mobile Actions accuracy increased from 58% to 85%), miniaturization for edge devices (such as NVIDIA Jetson Nano and mobile phones), and optimization for JSON/multilingual input. The official release provides download links via Hugging Face and Kaggle, as well as fine-tuning and deployment guides for Transformers, Unsloth, Keras, NeMo, and LiteRT‑LM, vLLM, MLX, Llama.cpp, Ollama, Vertex AI, and LM Studio. Multiple demos and datasets/Colab have also been launched on Edge Gallery.