Dyna Robotics releases Dyna-2.1 physical agent for long tasks
The robot runs whole-body workflows that last hours, built on the Dyna-2 world-action model.
Research#06
1M hours
Hours of human video used to pre-train Dyna-2
Image: Latent Space
1M hoursHours of human video used to pre-train Dyna-2
The short version
- Dyna-2 is a world-action model pre-trained on one million hours of human video.
- Dyna Robotics released Dyna-2.1, a physical agent that combines new semi-humanoid hardware with an agentic system built around Dyna-2.
- The company says Dyna-2.1 achieves reliable super long-horizon whole-body autonomy.
- Jason Ma said a Dyna model reached a 99.4% success rate folding restaurant napkins for 24 hours straight.
Why it matters
A world-action model predicts future video and uses that prediction to guide robot actions. The work points to long, reliable autonomy as a key problem for robots in real settings. For developers, it shows an agentic harness running multi-step physical workflows.
“So it’s the first world action model trained on more than one million hours of data.”