DiffusionGemma

Google has launched the new Gemma model under the Apache 2 license, following the earlier trial of the Gemini Diffusion model in May. The new model is available on NVIDIA's NIM cloud platform, facilitating access for developers. Gemma boasts improved performance, achieving a generation speed of up to 500 tokens per second, making it an attractive option for AI applications. The model's API has been utilized to generate content quickly and efficiently. This launch reflects Google's commitment to developing open AI models, expected to significantly impact the developer community and open new avenues for AI applications.