AI MODELS
Follow important model changes, understand what changed, and check the official evidence.
New versions, capabilities, access and real-world deployments.
UnifoLM-WLA
Unitree released the Base checkpoint and fine-tuning code.
Previously reported
September 20: model modules and action-expert training code available
This announcement
September 28: Base weights and fine-tuning code available
UnifoLM-WLA
Unitree adds model modules and action-expert training code.
Previously reported
September 11: UnifoLM-ER-1 and ER-Flow weights released
This announcement
September 20: model modules and action-expert training code released
Figure Helix
Figure reports tests across 30 unseen homes.
Previously reported
Helix 02: data collected in deployment environments
This announcement
Helix 2.5: three whole-body tasks in 30 unseen homes
Search models, developers or robots; explore versions, updates and official sources.
| Model series | Research organization / team | Type | Resources | Recorded versions | Version release date | Related robots | Actions |
|---|---|---|---|---|---|---|---|
| LingBot-VA Autoregressive video-action world model family for robot control. | Research organization / team | TypeWorld-action model | Resources | Recorded versionsLingBot-VA 2.0 | Version release dateJuly 9, 2026 | Related robots— | Actions Choose a version to compare |
| FastWAM World-action model family that learns future imagination but acts without rollout at inference. | Research organization / team Galaxea Dynamics 2 institutions | TypeRobot policyWorld-action model | Resources | Recorded versionsFastWAM base | Version release dateMarch 18, 2026 | Related robotsSO-101 | Actions |
| DreamZero NVIDIA world-action model jointly predicting video and robot actions. | Research organization / team | TypeWorld-action model | Resources | Recorded versionsDreamZero | Version release dateFebruary 17, 2026 | Related robots— | Actions |
| Poke-Aurora Poke-Aurora is PokeBot’s world-action model. RoboAPI accepts camera images, robot state and task instructions and returns actions. Joint video and action learning targets zero-shot generalization. Access is a noncommercial technical preview for eligible users. | Research organization / team | TypeWorld-action model | Resources— | Recorded versionsPoke-Aurora | Version release date— | Related robots— | Actions |
— means no confirmed information; it does not mean unsupported or unavailable.
