---
license: gemma
base_model: google/gemma-3-4b-it
pipeline_tag: text-generation
library_name: rkllm
tags:
- rkllm
- rockchip
- rk3576
- rk3588
- gemma
- gemma-3
---

# gemma-3-4b-it-RKLLM

RKLLM-converted Gemma 3 4B IT language-model artifacts for Rockchip RK3576 and RK3588 NPUs.

Important: the upstream [google/gemma-3-4b-it](https://huggingface.co/google/gemma-3-4b-it) model is multimodal/VLM-capable, but this repository contains only RKLLM language-model binaries. The converted model exposed here is LLM-only: it does not accept images and does not include an RKNN vision encoder.

These hardware-specific `.rkllm` files require a compatible Rockchip RKLLM runtime. They are not Transformers checkpoints and cannot be loaded directly with Transformers, llama.cpp, or Ollama.

## Base model

- Upstream model: [google/gemma-3-4b-it](https://huggingface.co/google/gemma-3-4b-it)
- License: Gemma Terms of Use
- Upstream type: VLM/multimodal
- RKLLM type: LLM-only

Review Google's Gemma terms before use or redistribution.

## Conversion and variants

### Toolkit version

**RKLLM Toolkit: v1.2.3**

Use a file built for the exact target SoC.

| Target | Quantization | File | SHA256 |
| --- | --- | --- | --- |
| RK3576 | W4A16 | [gemma-3-4b-it_RK3576_w4a16.rkllm](RK3576/gemma-3-4b-it_RK3576_w4a16.rkllm) | `0e6d6298453f483e832a7982bcdf1acdf895a47c6777e93eb3c1e43e82cf8fc2` |
| RK3576 | W8A8 | [gemma-3-4b-it_RK3576_w8a8.rkllm](RK3576/gemma-3-4b-it_RK3576_w8a8.rkllm) | `b077e9ee75f837230dbf1be948a03d39942abd90ab23343399fa13b7d19526cf` |
| RK3588 | W8A8 | [gemma-3-4b-it_RK3588_w8a8.rkllm](RK3588/gemma-3-4b-it_RK3588_w8a8.rkllm) | `aa04fce6ca2e418cd946814f1563fe8ed7e56dfbf5cf99ced4e753739287ee80` |

The repository also includes `gemma-3-4b-it_data_quant.json`, used as calibration data during conversion.

## Usage

```bash
hf download HanzoHuang/gemma-3-4b-it-RKLLM \
  RK3576/gemma-3-4b-it_RK3576_w4a16.rkllm \
  --local-dir gemma-3-4b-it-RKLLM
```

Run the file with the RKLLM runtime and the upstream Gemma instruction prompt format. For Docker deployment, see [Hanzo-Huang/rkllm-docker](https://github.com/Hanzo-Huang/rkllm-docker).

## Limitations

Vision input is not supported by these artifacts. Conversion can change output quality relative to the upstream model; validate the result on your target device.

## Acknowledgements

Thanks to Google, Rockchip, and the RKLLM community.