Google releases Gemma 3n — open-source multimodal model available

AI Impact Summary

Google has officially released Gemma 3n, a new open-source multimodal model, making it accessible across a wide range of popular open-source libraries including transformers, timm, MLX, llama.cpp, and Ollama. This release expands the model's utility by supporting image, text, audio, and video inputs, offering a versatile tool for developers and researchers. The availability of different model sizes (E2B and E4B) with varying memory requirements allows for deployment on a broader range of hardware, further enhancing its accessibility.