Google Enters Open-Source AI Race with New Gemma Models
Google launches Gemma, a new family of free open-source AI models that directly compete with offerings from Meta and AI startups. This move marks a strategic shift as Google hedges its bets beyond its proprietary Gemini models.
Google introduces Gemma, a new family of free open-source AI models designed to compete directly with similar offerings from Meta and various AI startups. The name serves as a nod to Google's more powerful, proprietary Gemini models, providing a lighter alternative for developers who do not require top-tier capabilities.
This release marks a significant strategic shift for Google, as the company previously remains firmly in the proprietary AI camp alongside OpenAI and Microsoft. By embracing open-source, Google acknowledges the growing popularity and improving performance of freely accessible models, effectively hedging its bets in an increasingly divided AI landscape.
Google DeepMind develops Gemma in response to programmers who prefer mixing open-source and proprietary models based on their specific needs. By offering both Gemma and Gemini on the same cloud platform, Google hopes to attract more customers who want to avoid the delays and errors associated with moving data across multiple cloud providers.