Gemma 4 26B A4B IT

JSON →
google llm
text

A 26-billion-parameter instruction-tuned model from Google's Gemma 4 family, using a 4-billion active parameter mixture-of-experts architecture.

context window 256K tokens
max output 128K tokens
input price $0.15 / 1M tokens
output price $0.6 / 1M tokens
code-generationreasoningstreamingopen-weightsfine-tunablefunction-callingtool-usevision
releasedApr 2025