Don't trust demos. Try it yourself.
Introducing RoboAPI, powered by Poke-Aurora, our world-action model.
Give compatible robots natural-language instructions through a simple API. No task-specific training.
More information: https://t.co/BHzTOo4I8u
Really enjoy playing with Poke’s RoboAPI in our lab. The whole system is plug-and-play. I’ll definitely stay tuned and look forward to their future updates.
@ClaraChengGo A robot needs more than the ability to follow instructions. It needs to understand how the physical world responds to its actions. That’s why we believe World-Action Models are such an important direction for embodied AI. 🤖
Thanks for the support! There’s definitely more to share about our evaluations and what we’re learning from real-world testing. If you have a UR or Franka arm, we’d love for you to give RoboAPI a try and share your experience!
Super Cool! Look forward to more sharing about it's eval.
Zeroshot on unseen object based on PokeBot WAM. UMI-style post-training and leverage wrist cam and standardized EEF control to adapt to different arms,
If you have Franka / UR arm, have a try!
@eastblood_tr That’s exactly why we made RoboAPI available as a technical preview. We don’t expect people to just take our word for it. Try it on your own hardware, bring your own tasks, and tell us what works and what doesn’t.
@eastblood_tr RoboAPI’yi tam da bu yüzden teknik önizleme olarak kullanıma açtık. Sadece bizim söylediklerimize güvenmenizi beklemiyoruz. Kendi donanımınızda deneyin, kendi görevlerinizi test edin ve nelerin işe yarayıp nelerin yaramadığını bize anlatın.
Don't trust demos. Try it yourself.
Introducing RoboAPI, powered by Poke-Aurora, our world-action model.
Give compatible robots natural-language instructions through a simple API. No task-specific training.
More information: https://t.co/BHzTOo4I8u
Thanks for helping more people discover RoboAPI! Our goal is to make experimenting with robot intelligence more accessible. Now it’s time to put it to the test! 🚀
PokeBot has launched RoboAPI, a public API that lets robot arms perform new tasks without collecting data or training a dedicated model.
Users can connect a supported arm with a wrist-mounted camera, give it a task such as "clean the table", and have RoboAPI send back its next moves in near real time.
#PokeBot
What if working with robots felt as easy as calling an LLM API?
That’s the idea behind RoboAPI. Instead of collecting data and training a new policy for every task, developers can simply send an instruction and let the model handle the actions.
Less time training. More time building. That’s the shift we’re excited about.
@rohanpaul_ai Making robot intelligence accessible through an API is just the starting point. The really interesting part will be seeing what developers build when they can experiment with physical AI without starting from scratch every time.
@ADLXBT@ZhihuFrontier That’s an exciting direction. The more easily robots can adapt to new tasks and environments, the closer we get to truly general-purpose physical intelligence.
We’ve seen language models become increasingly capable through APIs. Could robot intelligence follow a similar path?
The interesting question is what happens when developers can experiment with physical intelligence on real hardware, rather than just access another model endpoint.