MANUS™

Our research shows that cat videos are still more popular than haptic data glove videos. We were baffled by this result too. So we decided to combine both. Check it out and feel the purr-fection.

Physical AI needs data that text and video can't provide. @RLWRLD_ai is collecting it at the source: skilled workers, real tasks, real environments. MANUS gloves stream full hand and finger articulation into the training pipeline. Read the full use case: manus-meta.com/use-cases/how-… #robotics #humanoids #PhysicalAI #teleoperation #EmbodiedAI

Meet MANUS in Boston! Catch us at the @Robotics_Summit (May 27-28). High-precision data gloves for teleoperation & embodied AI training. Our Business Development Director Violaine Gragy will be on-site, let's connect. Learn more about MANUS in your robotics pipeline → manus-meta.com/robotics

Check out this video of @ManusMeta gloves teleoperating the 22-DoF @SharpaRobotics Wave Robotic Hand, capturing high-fidelity demonstration data for embodied AI. MANUS™ gloves provide high‑fidelity motion data for dexterity training and teleoperation. Their EMF‑based sensing system enables drift‑free, occlusion‑free tracking without the limitations of vision‑based approaches. Learn more about MANUS™ here: humanoidroboticstechnology.com/company/manus/ #humanoid #humanoids #humanoidrobotics #humanoidrobot

I promise this will be the best 20 min you spend today! Robotics: Endgame, the sequel to my last year's Sequoia AI Ascent talk, "Physical Turing Test". I laid out the roadmap for solving Physical AGI as a simple parallel to the LLM success story. Be a good scientist, copy homework ;) And stay till the end, more easter eggs and predictions for your polymarket! 00:30 DGX-1 origin story at OpenAI, I was there in 2016 signing with Jensen and Elon. Heading to the Computer History Museum! 01:42 The Great Parallel 03:31 Robotics, the Endgame 03:39 Why VLAs fall short 04:32 Video world models as the 2nd pretraining paradigm 06:09 World Action Models (WAM) 07:46 Strategies for robot data collection and the FSD equivalent to physical data flywheel for robot manipulation 11:06 EgoScale and the Dexterity Scaling Law we discovered recently 14:00 Physical RL: bridging the last mile 15:39 DreamDojo: an end-to-end neural physics engine for scaling RL in silico 17:00 Civilizational Technology Tree and my predictions for the near future. Spoiler: it's closer than you think. Thanks to my friends at Sequoia for inviting me back to AI Ascent this year! I had a blast! Last year's talk is attached in the thread if you missed it.

"Given his salary, I think this is by far the most expensive teleop trajectory ever collected in our dataset." Jim Fan (@DrJimFan) at @sequoia AI Ascent 2026, on the clip of NVIDIA Chief Scientist Bill Dally wearing MANUS gloves to teleoperate a humanoid robot inside the GEAR Lab. At @NVIDIARobotics, MANUS gloves provide high-fidelity hand motion data for the next generation of Physical AI models. Watch the full video: youtu.be/3Y8aq_ofEVs

MANUS gloves support multi-user capture out of the box. Every glove gets its own independent data stream, with no loss in precision.

N1 Robotics built Waldo, a humanoid teleoperation platform that uses MANUS gloves as the operator input layer. WaldoRT, their custom retargeter, maps MANUS data directly onto robot hands, preserving grasp intent across Inspire, BrainCo Revo 2, Sharpa, and others. As long as a URDF is available, WaldoRT generates the retargeting configuration for that hand, no per-operator or per-end-effector retuning required. Learn more: manus-meta.com/use-cases/how-… #robotics #teleoperation #datacollection #humanoids

@yavuzhankutuk @finkd @Meta You can check out this paper benchmarking MediaPipe-based monocular tracking vs our gloves on 30 dexterous manipulation tasks: gorgeous2002.github.io/TeleOpBench/

@ManusMeta @finkd @Meta bro mediapipe already do this without markerless and just via basic camera, so it looks like you discovered america again, well done;)

Hey Mark Zuckerberg @finkd and the @Meta team, sorry to hear the other Manus didn't work out.💔 If you're in the market for a rebound, we're Dutch, and we're available. Precision hand-tracking since 2014. DMs open.😉

MANUS gloves live at the 3rd China Embodied AI & Humanoid Robotics Conference in Beijing. Thanks to our partner CNBest for the on-site teleoperation demo! @Tesollo_Inc #robotics #teleoperation #EmbodiedAI #datacollection

@BlakeC @finkd @Meta Are you sure 😉

@ManusMeta @finkd @Meta Apple does this with cameras, no glove.

How humans move their hands is still the best blueprint for teaching robots to do the same. @mimicrobotics, a Switzerland-based physical AI company with roots in @ETH Zurich, builds dexterous robotic hands and the manipulation models that drive them. To train those models at scale, they use MANUS data gloves to capture human hand motion, turning real-world demonstrations into the datasets their manipulation policies learn from. #robotics #datacollection #EmbodiedAI

Tesollo, a South Korean company specializing in dexterous robotic hands, uses MANUS gloves to teleoperate its robotic hands in real time for precise paper cup stacking. MANUS gloves capture subtle hand movements with 25 degrees of freedom and millimeter-level precision without occlusion or drift. Every motion is mapped 1:1 to the robot end effector, enabling reliable teleoperation and high-quality data collection for dexterous manipulation. @Tesollo_Inc #robotics #teleoperation #datacollection #EmbodiedAI

Teleoperating the 22-DoF Sharpa Wave @SharpaRobotics with MANUS gloves, capturing high-fidelity demonstration data for embodied AI. #robotics #teleoperation #manipulation #datacollection

Our team just wrapped a packed trip across China. A few highlights: ✨ China Humanoid Robot Ecology Conference. MANUS gloves were running in demos across the show floor. A special shout-out to our partners at @roboterax. Their newly developed xhand1 Pro was teleoperated with MANUS at the booth, and the demo drew a serious crowd. ✨ Global Embodied Intelligence and Humanoid Innovation (GEIA Asia 2026). Great to connect with partners, customers, and peers across the APAC humanoid ecosystem. ✨ Humanoid Robot Training Webinar. We shared how MANUS hand-tracking is being adopted across teleoperation, data collection, and sim-to-real pipelines in humanoid development. ✨ @AGIBOTofficial's Dexterous Hand Developer Salon hosted by AGILINK. In-depth exchanges with the engineers and researchers building the next generation of manipulation, exactly the kind of conversations that move the field forward. ✨ And back-to-back customer meetings, every single day. The enthusiasm on the ground says everything about where this ecosystem is headed. China's humanoid teams are shipping fast. MANUS is proud to provide the precision and reliability their pipelines actually depend on.

The @manusmeta Metaglove controlling the Ability Hand wirelessly, in real time. What I keep coming back to isn't the tech. It's how natural the control feels. Intent captured, translated, executed - fluidity that makes you forget there's a signal chain in between. @PSYONICinc

Researchers from UC San Diego, UC Berkeley, and Amazon FAR (Frontier AI for Robotics) chose MANUS gloves to capture hand teleoperation data for humanoid robot training. The result: 82.5% task success rate on the Unitree G1, a 57% improvement over baseline. The precision of your data collection is a multiplier on downstream policy performance. Learn more: xl-vla.github.io Paper: arxiv.org/abs/2603.10158 #robotics #EmbodiedAI #humanoids #teleoperation #datacollection

Real-time teleoperation of the AGILINK Omnihand, powered by MANUS data gloves. AGILINK is the dexterous hand subsidiary of AGIBOT, one of China's leading humanoid robotics companies, developing next-generation robotic hands for real-world manipulation. MANUS gloves provide the high-fidelity hand tracking needed to drive the Omnihand in teleoperation workflows. @AGIBOTofficial #robotics #teleoperation #teleoperation #datacollection