Paimon-0 model release and benchmark performance
Overview
Primotion, a Shenzhen-based company established just three months ago, has launched its embodied base model Paimon-0, which has achieved first place in several benchmarks including RoboColiseum, RoboCasa365, and VLA-Arena. The model employs a World Action Model (WAM) and Inverse Dynamics Model (IDM) architecture, combined with FPGA hardware control to achieve low latency. The company enhances the model's generalization capabilities through data closed-loop and self-improvement systems (RSI), and plans to open-source the Progress Model Paimon-Eval.
From QbitAI — Robotics on WeChat
Event evidence and corrections
1 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · task_success_rate: 0.6613 other · Basis not reported · Differing source assertions
“要知道，这个榜最近可是刚把GPT-6 Astra、Claude Opus 5.5、Gemini 4 Argon这些通用多模态大模型请进场同台比试的： 虽说大模型在指令跟随和空间理解上分数确实亮眼，可一到要真刀真枪动手的通用操作子榜，Paimon-0的0.6613依然是全场最高，扰动适应子榜也由它领跑。”
Exact source · revision 2qbitai
Reported quantity · task_success_rate: 0.73 other · Basis not reported · Differing source assertions
“VLA-Arena 为通用具身模型开源评测榜单，围绕安全性、抗干扰能力做系统化测评，支持跨模型对比，Paimon-0取得0.73平均成功率，排名第一。”
Exact source · revision 2qbitai
Reported quantity · data_collection_cost: 500 other · Basis not reported
“单个任务靠后训练把成功率提上去，很多时候已经能满足客户要求，真正卡脖子的是数采，一个任务动辄要采500到1000条数据，实验室采完还得去客户现场接着采。”
Exact source · revision 2qbitai
Reported quantity · Reported duration: 2000 hours · Basis not reported
“RoboCasa365 是面向通用机器人策略的大规模多任务基准，包含2500+仿真厨房场景、365项日常任务、2000+小时交互数据，选取50项任务，从原子可见、组合可见、组合不可见三层检验泛化能力，Paimon-0以58.1的综合分数登顶。”
Exact source · revision 2qbitai
Reported quantity · tasks: 50 other · Basis not reported
“RoboCasa365 是面向通用机器人策略的大规模多任务基准，包含2500+仿真厨房场景、365项日常任务、2000+小时交互数据，选取50项任务，从原子可见、组合可见、组合不可见三层检验泛化能力，Paimon-0以58.1的综合分数登顶。”
Exact source · revision 1qbitai
Reported quantity · total_score: 46.599 other · Basis not reported · Differing source assertions
“同时Paimon-0参加启元机器人主办的 IROS Workshop 挑战赛，赛道聚焦双臂机器人策略，包含仿真线上评测与真机线下验证，炎曦动力最终以46.599总分夺冠，验证了端到端策略在双臂操作上的能力。”
Exact source · revision 1qbitai
Reported quantity · total_score: 58.1 other · Basis not reported · Differing source assertions
“RoboCasa365 是面向通用机器人策略的大规模多任务基准，包含2500+仿真厨房场景、365项日常任务、2000+小时交互数据，选取50项任务，从原子可见、组合可见、组合不可见三层检验泛化能力，Paimon-0以58.1的综合分数登顶。”
Exact source · revision 1qbitai
Evidence dependency changes (5)
- 2026-10-30T20:05:35.239Z · evidence updated · source revision 1. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-30T20:05:17.545Z · evidence updated · source revision 1. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-30T20:03:51.912Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-30T20:03:44.882Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-26T17:45:25.920Z · evidence updated · source revision 1. Evidence extraction was updated; current source attributions are shown above.
Report timeline
Follow attributed reports and material updates.
- QbitAI — RoboticsPrimotion Releases Embodied Base Model Paimon-0, Tops Multiple Embodied Intelligence Rankings
Shenzhen startup Primotion launched its embodied base model Paimon-0 within three months of its establishment, achieving top rankings in several benchmarks including RoboColiseum, RoboCasa365, and VLA-Arena.
- QbitAI — Robotics on WeChatPaimon-0 Leads in Multiple Embodied Benchmark Tests, Shenzhen Startup Primotion Makes Debut
Primotion, a Shenzhen-based company established just three months ago, has launched its embodied base model Paimon-0, which has achieved first place in several benchmarks including RoboColiseum, RoboCasa365, and VLA-Arena. The model employs a World Action Model (WAM) and Inverse Dynamics Model (IDM) architecture, combined with FPGA hardware control to achieve low latency. The company enhances the model's generalization capabilities through data closed-loop and self-improvement systems (RSI), and plans to open-source the Progress Model Paimon-Eval.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).