Ant Group’s Robbyant Unveils LingBot-VA 2.0: A Causal Video-Action Model Built Natively for Physical AI
Ant Group’s Robbyant released LingBot-VA 2.0, a causal video-action model designed for physical AI systems. It predicts future states at 225 Hz for asynchronous control.
- LingBot-VA 2.0 is a Physical AI foundation model built natively for embodiment, not fine-tuned from video generators.
- The model achieves 225 Hz asynchronous control using a causal Diffusion Transformer and sparse Mixture-of-Experts video stream.
- Foresight Reasoning enables the model to predict future states ahead of execution and re-ground on real observations.
- The paper acknowledges discrepancies in its own reported metrics, indicating potential areas for further validation.
Ant Group’s Robbyant has introduced LingBot-VA 2.0, a new foundation model for Physical AI that departs from traditional video-action approaches. Unlike models fine-tuned from video generators, LingBot-VA 2.0 is built from scratch to support real-time embodiment in robots and autonomous systems. The model leverages a causal Diffusion Transformer (DiT) architecture and a sparse Mixture-of-Experts (MoE) video stream to achieve high-frequency asynchronous control at 225 Hz.
The technical report highlights a semantic visual-action tokenizer that converts raw video inputs into actionable representations, enabling the model to predict future states ahead of execution through a process called Foresight Reasoning. This capability allows the system to re-ground its predictions on every real observation, improving adaptability in dynamic environments. However, the paper also notes discrepancies in its own reported metrics, which may require further validation.
The release underscores Ant Group’s push into Physical AI, positioning LingBot-VA 2.0 as a tool for developers working on robotics, autonomous vehicles, and industrial automation. The model’s native design for embodiment could reduce the need for extensive fine-tuning, potentially accelerating deployment in real-world applications.
Provides a new foundation model for real-time physical AI control, reducing the need for fine-tuning.
Enables faster deployment of robotics and automation solutions with high-frequency control capabilities.
Signals Ant Group’s strategic expansion into Physical AI, a growing segment with high commercial potential.
Demonstrates advancements in AI models designed for real-world physical interactions.
- Foresight Reasoning
- A process where the model predicts future states ahead of execution to improve real-time decision-making.
- Diffusion Transformer (DiT)
- A neural network architecture combining diffusion models with transformer layers for generative tasks.
- Mixture-of-Experts (MoE)
- A model architecture where multiple specialized sub-models (experts) are combined dynamically for improved performance.
Mistral joins rush to develop AI for robots - InfoWorld
RoboticsHumanoid robots controlled by surgeons did world-first operation on live pigs
RoboticsThe 1X Neo Robot Has Freaky Fast Fingers
RoboticsRobbyant Releases LingBot-VLA 2.0: An Open-Source 6B Vision-Language-Action (VLA) Model for Cross-Embodiment Robot Manipulation
RoboticsRobbyant Releases LingBot-VLA 2.0: An Open-Source 6B Vision-Language-Action (VLA) Model for Cross-Embodiment Robot Manipulation
AI ResearchChina's Orca world model matches specialized robotics systems without ever seeing a single action label
China’s Beijing Academy of Artificial Intelligence unveiled Orca, a world model trained on 125,000 hours of video that predicts abstract states instead of tokens or pixels. It matches specialized robotics systems on five benchmark tasks without using any action labels.
AI moved in next door. For this Memphis community, life got more complicated. - The Christian Science Monitor
A Memphis community is grappling with the arrival of AI, leading to increased complexity in daily life. The Christian Science Monitor reports on the challenges faced by residents.
Macroscope | Don’t expect the rising tide of AI to lift all boats - South China Morning Post
A recent article by the South China Morning Post warns that the growing use of AI may not have a positive impact on all industries.
LLMMeta's Muse Spark 1.1 outperforms GLM-5.2 in coding and costs slightly less
Meta’s Muse Spark 1.1 improves coding performance by 8 points on the Artificial Analysis Intelligence Index and reduces hallucination rates by 35 percentage points while costing less than GLM-5.2.
BusinessOpenAI admits it "didn't get everything quite right" with ChatGPT Work launch and scrambles to fix UX and costs
OpenAI has admitted to multiple issues with its new ChatGPT Work and GPT-5.6 Sol releases, including excessive compute usage, confusing desktop transitions, and unauthorized data deletions.
BusinessApple sues OpenAI for allegedly running a "coordinated campaign" to steal trade secrets through poached employees
Apple has filed a lawsuit accusing OpenAI of orchestrating a campaign to poach employees and steal trade secrets related to unreleased products.