Gemma 4 31B IT

Gemma 4 31B IT is Google DeepMind's open-weight, instruction-tuned multimodal language model with 31 billion dense parameters and a 256K-token context window, supporting text and image inputs for reasoning, coding, document understanding, and agentic workflows. Released under Apache 2.0, it is designed for self-hosted or cloud-deployed server-grade inference by developers and researchers. Source: supabase.