What stands out to me about Liber-0 is the attempt to make human experience useful to robots, not just collect more data.
It was tested on 60 tasks (600 trials), with a 36% overall success rate: 63.5% on pick-and-place, 34% on non-PnP tasks, and 10.5% on the more challenging multi-step tasks.
That spread is telling: zero-shot transfer is promising, and the hardest tasks are still hard.
The combination of glove-grounded actions and embodiment alignment is what I’ll be watching. Curious to see how this develops.
Introducing Liber-0, our World Action Model for zero-shot robot manipulation—including dexterous, multi-step tasks.
Across 60 tasks proposed by GPT-6 and 600 trials, Liber-0 achieved 36% overall success without robot demonstrations collected specifically for those tasks:
• Pick-and-place: 63.5%
• Other manipulation tasks: 34.0%
• Challenging multi-step tasks: 10.5%
Zero-shot capability extends beyond individual actions to coordinated sequences.
Pretrained on 2.2M+ hours of human experience, Liber-0 jointly predicts future observations and actions to turn human experience into robot skills.
Technical blog:
Meet PrimeBOT T1 🤖 Price starts at RMB 19,999 (approx. US$2,960), Overseas pricing TBA.
They are scaling up and producing one humanoid robot off the line every 2.5 minutes, with capacity for 10,000 per month.
AgiBot is inviting channel partners to an exclusive, invitation-only summit in Santa Clara on October 15.
Experience live demos of our latest robots, learn about channel partner policy, and explore new opportunities.
Scan the QR code or apply here:
I’m excited to announce that my online shop is now live.
A curated collection of robotic hardware, now available with simple, all-in pricing.
No hidden costs. No extra domestic shipping fees.
Orders ship domestically from the U.S. with short lead times.
Explore the shop:
100M ARR, 60+ paid customers, 10% comes from mobility and the rest from manipulation related deployments.
Welcome to the mass deployment phase of robotics.
We have a huge news to share today!
Today we are unveiling the first truly accessible RL robot - welcome Microduck
A 25 cm tiny open-source biped with 15 actuators and packed with sensors (camera, speaker, LiDAR, NFC, bluetooth, wifi, etc) that you train yourself with reinforcement learning.
It's also playable out of the box with more than half a dozen fun and playful pre-trained policies to have it walk, sit, crouch, roller-skate, pick up objects with its articulated beak, and recover on its own.
And all for less than $400.
See all the details, play with the simulator and order it at:
(video with sound on 🔊)
GRIT, a humanoid robotics lab in China demos its whole-body control under strong terrain disturbances during teleoperation.
The robot keeps tracking the operator while stepping over obstacles and maintaining balance.
Pretty robust.