π0 與 π0-FAST：用於通用機器人控制的視覺-語言-動作（VLA）模型
Original: π0 and π0-FAST: Vision-Language-Action Models for General Robot Control
Physical Intelligence, a Physical AI startup founded by robotics luminary Sergey Levine and others, has officially open-sourced its…
機器人 AI 新創公司 Physical Intelligence 在 Hugging Face 上開源了其通用機器人基礎模型 π0 及其加速版本 π0-FAST。這款視覺-語言-動作（VLA）模型能透過文字與視覺指令控制多種不同硬體的機器人，執行摺衣服、收拾雜物等複雜任務。π0-FAST 則大幅提升了推理速度，滿足高頻率實時控制的需求，為開源實體 AI 領域帶來重大突破。
Physical Intelligence, a Physical AI startup founded by robotics luminary Sergey Levine and others, has officially open-sourced its flagship robot foundation models π0 (Pi0) and the high-frame-rate variant π0-FAST on Hugging Face. This marks a major step forward for general-purpose robot control technology toward the open-source community.
Free shows the 3-line summary; Pro unlocks the full deep summary (~300 words) so you never have to click through.
See Pro plans →Want the original English / full article?
Read on Hugging Face Blog →Summaries are AI-generated; the original article is authoritative.