MiniCPM-V 4.6 as a VLM backbone in StarVLA for LIBERO robot manipulation

Hi MiniCPM-V team,

I wanted to share a downstream integration / community use case of MiniCPM-V 4.6 in embodied AI.

We recently integrated MiniCPM-V 4.6 as a new vision-language backbone for StarVLA, a Vision-Language-Action framework for robot manipulation.

StarVLA PR: https://github.com/starVLA/starVLA/pull/354

The integration adds:

• A MiniCPM-V 4.6 VLM wrapper for StarVLA
• PI / GR00T-style VLA framework support
• LIBERO training and evaluation scripts
• Example configs for 8-GPU training

Using MiniCPM-V 4.6, we trained and evaluated on the LIBERO benchmark and got the following results:

Benchmark	Success Rate
LIBERO-Spatial	94.0%
LIBERO-Object	98.0%
LIBERO-Goal	98.0%
LIBERO-10	92.4%
Overall	95.6%

Training setup:

• Backbone: openbmb/MiniCPM-V-4.6
• Effective batch size: 128
• Max training steps: 80k
• Attention implementation: flash_attention_2
• No modules are frozen by default (FREEZE_MODULES=""), so MiniCPM-V is trainable unless overridden.

MiniCPM-V 4.6 looks like a promising lightweight VLM backbone for embodied AI / VLA-style robot learning. The initial LIBERO results are encouraging, especially given the compact model size compared with many larger VLM backbones.

As a next step, we are also considering testing this MiniCPM-V-based StarVLA model on a real SO-101 robot setup, to further evaluate its sim-to-real potential beyond LIBERO simulation benchmarks.

Thanks for releasing MiniCPM-V!