Google’s Gemma AI models hit 150 million downloads since launch

Google’s Gemma AI models have reached over 150 million downloads since their launch in February 2024. This milestone was announced by Omar Sanseviero, a developer relations engineer at Google DeepMind.

Developers have also created more than 70,000 variants of Gemma on the AI development platform Hugging Face.

This achievement marks a significant point in Google’s AI model adoption and community engagement.

Widespread Adoption and Significance

Gemma models are designed to be lightweight and accessible for various devices, from laptops to cloud systems. This design aims to democratize access to advanced AI capabilities across many user groups.

The rapid adoption highlights a larger trend toward open-source AI models that emphasize ethics and performance. Google’s commitment to responsible AI is evident in the thorough testing and validation of Gemma models.

These models work with both images and text, supporting over 100 languages for broader usability.

Responsible Development and Challenges

Google released the Responsible Generative AI Toolkit to help developers deploy Gemma responsibly and ethically. This toolkit guides users in avoiding biases and harmful content generation while using the models.

Gemma has specialized versions fine-tuned for tasks such as drug discovery and emotion recognition.

Despite its success, Gemma trails Meta’s Llama models, which surpassed 1.2 billion downloads in April 2025.

Both Gemma and Llama face criticism for non-standard licensing terms, creating potential commercial usage risks.

Impact on Google and the AI Sector

This milestone reinforces Google’s influence in AI development and open-source model availability. Alphabet Inc., Google’s parent company, maintains strong investor confidence with a market value of around $1.88 trillion.

Gemma’s success reflects growing demand for accessible and ethically designed AI tools worldwide.

Further development of Gemma models aims to push innovation in the responsible AI space.

Developers and users are invited to suggest improvements for the next Gemma versions.