Gemma 4 26B
Gemma 4 26B A4B by Google DeepMind is an open-weight multimodal mixture-of-experts model built for reasoning, coding, and agentic workflows. It combines 25.2B total parameters with about 3.8B active parameters per token, supports up to 256K context, and accepts text and image input with configurable thinking mode, native function calling, and system-role support. Official Gemma 4 benchmarks position it especially well for math reasoning, coding, multilingual tasks, and multimodal understanding, making it a strong fit for coding assistants, tool-using agents, long-context knowledge work, and document or image-grounded workflows.
Gemma 4 26B
Gemma 4 26B A4B by Google DeepMind is an open-weight multimodal mixture-of-experts model built for reasoning, coding, and agentic workflows. It combines 25.2B total parameters with about 3.8B active parameters per token, supports up to 256K context, and accepts text and image input with configurable thinking mode, native function calling, and system-role support. Official Gemma 4 benchmarks position it especially well for math reasoning, coding, multilingual tasks, and multimodal understanding, making it a strong fit for coding assistants, tool-using agents, long-context knowledge work, and document or image-grounded workflows.
