MiniCPM4-0.5B-RKLLM

RKLLM-converted MiniCPM4-0.5B language-model artifacts for Rockchip RK3576 and RK3588 NPUs.

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

Base model

• Upstream model: openbmb/MiniCPM4-0.5B
• License: Apache-2.0
• Model type: LLM (text only)

Conversion and variants

Toolkit version

RKLLM Toolkit: v1.3.0

Use a file built for the exact target SoC.

Target	Quantization	File	SHA256
RK3576	W4A16 (g128)	MiniCPM4-0.5B_RK3576_w4a16_g128.rkllm	34060c78f36ee6ac2a772d5353d32d700fe940ea48564f083b8665a85fdc499c
RK3576	W8A8	MiniCPM4-0.5B_RK3576_w8a8.rkllm	10f2bfa298b5a79077fc52f339ead0446e9e834632226a8bfcc3d003d42a0ba8
RK3588	W8A8	MiniCPM4-0.5B_RK3588_w8a8.rkllm	fef1e215dfa69391d2c5da53eb918020419a71e6b3539d9ef91d56d956b602ea

The repository also includes MiniCPM4-0.5B_data_quant.json, used as calibration data during conversion.

Usage

Download a matching artifact with the Hugging Face CLI:

hf download HanzoHuang/MiniCPM4-0.5B-RKLLM \
  RK3576/MiniCPM4-0.5B_RK3576_w4a16_g128.rkllm \
  --local-dir MiniCPM4-0.5B-RKLLM

Run it with the RKLLM runtime and the upstream MiniCPM4 chat template. For a Docker deployment, see Hanzo-Huang/rkllm-docker.

Limitations

Conversion can change output quality relative to the upstream model. Runtime, driver, memory, and toolkit compatibility depend on the Rockchip software stack installed on the device.

Acknowledgements

Thanks to OpenBMB, Rockchip, and the RKLLM community.