# XPENG ROBOTICS releases embodied VLM Capek 0.5: 2B and 35B-A3B sizes, improving on 30 and 28 of 34 benchmarks respectively

- Source: 3D视觉工坊 · 微信公众号
- Published: 2026-11-03T01:30:00.000Z
- RoboHot: https://robohot.news/en/items/peueyrj4vnzml8r1z6ocyutis
- Original: https://mp.weixin.qq.com/s?__biz=MzU1MjY4MTA1MQ==&mid=2247752821&idx=2&sn=a1415323dae1519aa22f262bfaa5c7e7&chksm=fa647748ad0e01ead6831cb49f087617a5cf037388b502dfcabf6333751976a8faf438cf91e6&scene=126&sessionid=1793703610#rd

## Summary

XPENG ROBOTICS proposes the embodied vision-language model Capek 0.5, training separate experts for four execution capability categories—spatial reasoning, temporal understanding, action guidance, and state verification—then integrating them into a single model via TIES merging and routed MOPD distillation.