GoogleGemma 4
Gemma 4 12B Unified
Gemma 4's encoder-free multimodal checkpoint.
What the release provides
Its 11.95B parameters process text, images and audio without separate modality encoders. Context is 256K; the license is Apache 2.0.Google Gemma 4 model card
How to use it well
Practical guidance. Use the complete multimodal serving path when reviewing this checkpoint. Check audio, image and text processing together, including how inputs are prepared. A text-only timing result cannot establish the latency or memory needs of a multimodal application.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | google/gemma-4-12B-it |