Together AI Expands Model Portfolio With Integration of Gemma 4 31B

According to a recent LinkedIn post from Together AI, the company is making Google DeepMind’s Gemma 4 31B model available on its infrastructure. The post highlights capabilities such as multimodal reasoning with text and images, document parsing, optical character recognition, and native function calling for tool-oriented agent workflows.

As shared in the post, Together AI positions this integration as production-ready on its AI Native Cloud, citing a 99.9% service level, 256K context window, and support for over 140 languages. For investors, this suggests an effort to deepen the platform’s value proposition for AI-native customers, which could enhance usage-based revenue potential and strengthen Together AI’s competitive stance in the rapidly evolving inference and agentic AI infrastructure market.