Runway releases open-weight Praxis-1 robot control model
Start here
Runway has released Praxis-1, a robot control model built on video pretraining, with its weights and source code published under the Apache 2.0 license, which permits commercial use. Developers can download it from Hugging Face after signing up and accepting the access terms. The model takes camera images, text instructions and robot position data, predicts how the scene will change, and derives matching robot movements. Runway reports a 46.2 percent success rate on Nvidia's RoboLab-120 leaderboard, ahead of FLUX 3 Action at 42.9 percent; Runway ran all the tests itself and the results have not been independently verified. The standard version needs an 80 GB GPU and further training before it works with new robot systems. Runway's memory estimates suggest a compressed version should fit on a 32 GB RTX 5090, though the GitHub docs say actual peak memory use has not been measured.
Generated from reports · updated 3 h ago
Report timeline
Follow the reports to see the event from different sides.
- The DecoderRunway releases open-source Praxis-1 robot control model
Runway has released Praxis-1, a robot control model built on video pretraining, with open weights and source code under the Apache 2.0 license. The model takes camera images, text instructions and robot position data, predicts how the scene will change and derives matching robot movements, and it scored a 46.2 percent success rate on Nvidia's RoboLab-120 leaderboard, ahead of FLUX 3 Action at 42.9 percent. Runway ran all tests itself and the results have not been independently verified; the base model needs an 80 GB GPU and further training before working with new robot systems.
Heat for this event
Not enough continuous observations yet to draw a trend.