VLA Models
Vision-language-action models that map camera and language input directly to robot motion — architectures, benchmarks, and real-world deployments.
Coverage (17 articles)
- NVIDIA GR00T 1.7 and Agile ONE SN Launch at GTC Berlin — NVIDIA released GR00T 1.7, a 3-billion-parameter VLA model, and partnered with Agile Robots on the Agile ONE SN humanoid reference platform at GTC Berlin on Oct
- NVIDIA Eyes $1B Figure AI Stake at $38B Valuation — NVIDIA is in talks to invest another $1 billion in Figure AI at a roughly $38 billion pre-money valuation. The move would expand NVIDIA's existing stake and sup
- NVIDIA GR00T N2 Preview at GTC 2026 Advances Physical AI — NVIDIA previewed GR00T N2 at GTC 2026, claiming more than double the success rate on novel tasks compared to prior VLA models. The announcement included Cosmos
- Google DeepMind Gemini Robotics 2 Scales via Software Partnerships — Google DeepMind released Gemini Robotics 2 on July 30, 2026, delivering vision-language-action models for whole-body humanoid control, dexterity, and multi-robo
- NVIDIA Expands Open Physical AI Stack with OpenUSD for Humanoids — NVIDIA released updates to its open physical AI stack on September 30, 2026, built around OpenUSD and Omniverse for reusable digital twins in humanoid training
- XPENG IRON Humanoid Shows Memory and Multilingual Showroom Skills — XPENG released a September 22 video of its IRON humanoid in a lab test simulating car dealership interactions. The robot tracked speakers, remembered customer p
- Skild AI S1 Drives $100M Revenue Run Rate Ten Months In — Skild AI reached a $100 million annual recurring revenue run rate just ten months after its first commercial deployments of the S1 robotic foundation model. The
- NVIDIA SONIC Scales Humanoid Control Beyond Task-Specific Limits — NVIDIA researchers released SONIC, a single controller trained on over 100 million motion frames that lets humanoids handle diverse whole-body movements from VR
- Hugging Face LeRobot GR00T N1.7 Integration and Microduck Biped Launch — Hugging Face updated its LeRobot library in July 2026 to support NVIDIA's GR00T N1.7 VLA model for fine-tuning. In late August, its Pollen Robotics team opened
- NVIDIA Isaac GR00T N1.7 VLA Model Ships with Human Video Pretraining — NVIDIA has released the Isaac GR00T N1.7 VLA foundation model under Apache 2.0. The update adds pretraining on 32,000 hours of human video plus 8,000 hours of s
- Google DeepMind Launches Gemini Robotics ER 2 Embodied Reasoning Model — Google DeepMind released Gemini Robotics ER 2 on July 30, 2026. The embodied reasoning model enables robots to handle real-time spatial reasoning, multi-step pl
- Pi0.7 and Helix: VLA Transformers Advance Zero-Shot Dexterous Manipulation — Physical Intelligence's π₀.7 (April 2026) and Figure AI's Helix (February 2025) represent state-of-the-art Vision-Language-Action (VLA) models, achieving open-w
- ThorArena Benchmark Exposes Force-Aware Gaps in Humanoid Controllers — ThorArena, a new benchmark from TU Munich, tests how humanoid robots handle physical contact by replaying real human forces in simulation. The results expose ma
- Apptronik Robot Park Powers Apollo 3 Humanoid Data Factory — Apptronik's 90,000 sq ft Austin Robot Park, opened June 30 2026, deploys Apollo 2 fleets to generate real-world datasets for Gemini Robotics VLA models, acceler
- AI2 Robotics $735M Round Powers AlphaBot Wheeled Humanoid Toward 2026 IPO — Shenzhen's AI2 Robotics closed a $735M round at nearly $3B valuation for its AlphaBot wheeled humanoid with embodied AI, securing industrial orders like 1,000 u
- Skild AI's $1.4B Skild Brain: Omni-Bodied VLA Foundation Model — Skild AI closed a $1.4 billion Series C at over $14 billion valuation, led by SoftBank with NVIDIA's NVentures and Bezos Expeditions participating. The funding
- OpenVLA 7B Advances Zero-Shot VLA Control Over RT-2 — OpenVLA's 7B-parameter Transformer delivers superior zero-shot generalization on LIBERO benchmarks compared to RT-2-X while enabling high-frequency neural feedb
Related robot platforms
- Figure 03 — Figure AI · Humanoid