| OpenWAM: An Open, Modular Exploration Towards Systematic World-Action Model Pretraining | ✓ Link | 99.8 | OpenWAM-alpha (extra data) | 2026-09-07 |
| StreamPI: Streaming Multimodal Temporal Modeling for Vision-Language-Action Models | ✓ Link | 99.6 | StreamPI (T=5) | 2026-08-26 |
| Magic-W0: A Structured World-Action Foundation Model for Physical Intelligence | ✓ Link | 99.6 | Magic-W0 (extra data) | 2026-10-03 |
| Hume: Introducing System-2 Thinking in Visual-Language-Action Model | ✓ Link | 99.4 | Hume | 2025-05-27 |
| ACoT-VLA: Action Chain-of-Thought for Vision-Language-Action Models | ✓ Link | 99.4 | ACoT-VLA (LIBERO-Goal) | 2026-01-16 |
| Being-H0.5: Scaling Human-Centric Robot Learning for Cross-Embodiment Generalization | ✓ Link | 99.4 | Being-H0.5 (specialist) (LIBERO-Goal) | 2026-01-19 |
| BayesianVLA: Bayesian Decomposition of Vision Language Action Models via Latent Action Queries | ✓ Link | 99.4 | LangForce (LIBERO-Goal) | 2026-01-21 |
| PhysBrain 1.0 Technical Report | ✓ Link | 99.4 | PhysBrain 1.0 | 2026-05-14 |
| FM-Steer: Enhance Generalist Policies with Value-Guided Cascaded Denoising | | 99.4 | FM-Steer (π0) | 2026-06-03 |
| ABot-M0.5: Unified Mobility-and-Manipulation World Action Model | | 99.4 | ABot-M0.5 | 2026-07-01 |
| SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning | ✓ Link | 99.2 | SimpleVLA-RL (OpenVLA-OFT) | 2025-09-11 |
| HAMLET: Switch your Vision-Language-Action Model into a History-Aware Policy | | 99.2 | HAMLET | 2025-10-01 |
| Interactive Post-Training for Vision-Language-Action Models | ✓ Link | 99.0 | OpenVLA-OFT + RIPT | 2025-05-22 |
| Being-H0.5: Scaling Human-Centric Robot Learning for Cross-Embodiment Generalization | ✓ Link | 99.0 | Being-H0.5 (generalist) (LIBERO-Goal) | 2026-01-19 |
| ABot-M0: VLA Foundation Model for Robotic Manipulation with Action Manifold Learning | ✓ Link | 99.0 | ABot-M0 | 2026-02-11 |
| SnapFlow: One-Step Action Generation for Flow-Matching VLAs via Progressive Self-Distillation | | 99.0 | π0.5 + SnapFlow (1-step) | 2026-04-07 |
| Flex-$π$: A Multi-Stream World-Action Model with Compute Flexibility | ✓ Link | 99.0 | Flex-π* (full joint) | 2026-08-11 |
| Self-Improving Vision-Language-Action Models with Data Generation via Residual RL | | 98.9 | OpenVLA w/ PLD | 2025-10-30 |
| MorphAct: Robot Policies Morphing with the Worlds They Act In | ✓ Link | 98.87 | MorphAct (pi0.5) | 2026-10-14 |
| Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Model | ✓ Link | 98.8 | Spatial Forcing | 2025-10-14 |
| Mixture of Horizons in Action Chunking | ✓ Link | 98.8 | π0.5 with MoH | 2025-11-24 |
| LAP: Language-Action Pre-Training Enables Zero-shot Cross-Embodiment Transfer | | 98.8 | LAP-3B | 2026-02-11 |
| Xiaomi-Robotics-0: An Open-Sourced Vision-Language-Action Model with Real-Time Execution | ✓ Link | 98.8 | Xiaomi-Robotics-0 | 2026-02-13 |
| StreamingVLA: Streaming Vision-Language-Action Model with Action Flow Matching and Adaptive Early Observation | | 98.8 | StreamingVLA (AFM) | 2026-03-30 |
| ImageWAM: Do World Action Models Really Need Video Generation, or Just Image Editing? | ✓ Link | 98.8 | ImageWAM (FLUX.2 4B) | 2026-06-17 |
| PolicyTrim: Boosting Intrinsic Policy Efficiency of Vision-Language-Action Models | ✓ Link | 98.8 | π0.5 + PolicyTrim | 2026-06-21 |
| SimVLA: A Simple VLA Baseline for Robotic Manipulation | ✓ Link | 98.6 | SimVLA (LIBERO-Goal) | 2026-02-20 |
| StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developing | ✓ Link | 98.6 | StarVLA (generalist) | 2026-04-06 |
| InternVLA-A1.5: Unifying Understanding, Latent Foresight, and Action for Compositional Generalization | ✓ Link | 98.6 | InternVLA-A1.5-Libero | 2026-07-06 |
| When to Switch: Reliable Action-Chunk Extension for Vision-Language-Action Models | ✓ Link | 98.6 | RACE (H_exec=10) | 2026-10-05 |
| StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developing | ✓ Link | 98.5 | StarVLA-OFT (specialist) | 2026-04-06 |
| World-Language-Action Model for Unified World Modeling, Language Reasoning, and Action Synthesis | ✓ Link | 98.4 | WLA-0 +TTS (LIBERO-Goal) | 2026-06-04 |
| LaWAM: Latent World Action Models for Efficient Dynamics-Aware Robot Policies | ✓ Link | 98.4 | LaWAM (LIBERO-SFT) | 2026-06-14 |
| EmbodiedSkills: A Unified Framework for Orchestrating, Training, and Deploying VLA Agents | ✓ Link | 98.4 | EmbodiedSkills task-adapted pi0.5 | 2026-09-01 |
| VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Model | ✓ Link | 98.2 | VLA-Adapter-Pro (LIBERO-Goal) | 2025-09-11 |
| The Better You Learn, the Smarter You Prune: Towards Efficient Vision-Language-Action Models Via Differentiable Token Pruning | | 98.2 | LightVLA (LIBERO-Goal) | 2025-09-16 |
| Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning | | 98.2 | Cosmos Policy | 2026-01-22 |
| π-StepNFT: Wider Space Needs Finer Steps in Online RL for Flow-based VLAs | ✓ Link | 98.2 | π-StepNFT (π0.5, LIBERO-Goal) | 2026-03-02 |
| Fast-WAM: Do World Action Models Need Test-time Future Imagination? | ✓ Link | 98.2 | Fast-WAM-Joint | 2026-03-17 |
| FASTER: Rethinking Real-Time Flow VLAs | ✓ Link | 98.2 | X-VLA + FASTER | 2026-03-19 |
| Mimic Intent, Not Just Trajectories | ✓ Link | 98.2 | MINT-4B | 2026-03-28 |
| FrameSkip: Learning from Fewer but More Informative Frames in VLA Training | ✓ Link | 98.2 | FrameSkip | 2026-05-13 |
| IntentVLA: Short-Horizon Intent Modeling for Aliased Robot Manipulation | ✓ Link | 98.1 | IntentVLA | 2026-05-14 |
| MMaDA-VLA: Large Diffusion Vision-Language-Action Model with Unified Multi-Modal Instruction and Generation | ✓ Link | 98.0 | MMaDA-VLA | 2026-03-26 |
| StableVLA: Towards Robust Vision-Language-Action Models without Extra Data | | 98.0 | StableVLA (0.5B, clean) | 2026-05-18 |
| Patch Policy: Efficient Embodied Control via Dense Visual Representations | ✓ Link | 98 | Patch Policy + Diffusion Policy (WebSSL) | 2026-07-20 |
| SimpleMemVLA: A Simple but Effective Native-Video Memory for Vision-Language-Action Models | ✓ Link | 98 | SimpleMemVLA | 2026-09-02 |
| Conditional Trajectory Peaks: Single-Pass Multimodal Policies over Action Chunks | | 98.0 | CTP | 2026-10-05 |
| Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success | ✓ Link | 97.9 | OpenVLA-OFT (additional inputs) | 2025-02-27 |
| X-VLA: Soft-Prompted Transformer as Scalable Cross-Embodiment Vision-Language-Action Model | ✓ Link | 97.8 | X-VLA-0.9B | 2025-10-11 |
| Light-WAM: Efficient World Action Models with State-Fusion Action Decoding | ✓ Link | 97.8 | Light-WAM | 2026-06-06 |
| Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model | ✓ Link | 97.8 | Dynin-Robotics | 2026-09-11 |
| Fast-dVLA: Accelerating Discrete Diffusion VLA to Real-Time Performance | | 97.6 | Fast-dVLA | 2026-03-26 |
| Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models | ✓ Link | 97.6 | Embodied-R1.5-VLA | 2026-06-09 |
| LEAP-VLA: Latent-Enhanced Action Prototyping via Continuous Residual Latent Spaces for Vision-Language-Action Models | | 97.6 | LEAP-VLA | 2026-09-08 |
| GR00T N1: An Open Foundation Model for Generalist Humanoid Robots | ✓ Link | 97.5 | GR00T N1.7 | 2025-03-18 |
| VLA Knows Its Limits | | 97.5 | AutoHorizon (ω0.5, p=10) | 2026-09-08 |
| SwiftVLA: Unlocking Spatiotemporal Dynamics for Lightweight VLA Models at Minimal Overhead | ✓ Link | 97.4 | SwiftVLA with 4D input | 2025-11-30 |
| HiF-VLA: Hindsight, Insight and Foresight through Motion Representation for Vision-Language-Action Models | ✓ Link | 97.4 | HiF-VLA | 2025-12-10 |
| OxyGen: Unified KV Cache Management for Vision-Language-Action Models under Multi-Task Parallelism | ✓ Link | 97.4 | OxyGen (π0.5-LIBERO checkpoint) | 2026-03-15 |
| TBD-VLA: Temporal Block Diffusion Vision Language Action Model | | 97.4 | TBD-VLA | 2026-06-05 |
| Geometric Action Model for Robot Policy Learning | ✓ Link | 97.4 | GAM | 2026-06-15 |
| TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with <1 GB VRAM | ✓ Link | 97.4 | TurboVLA (DINOv3 ViT-B) | 2026-07-29 |
| BridgeWAM: Bridging Video and Action Experts via Latent Bridge Queries | ✓ Link | 97.4 | BridgeWAM | 2026-10-10 |
| VLASH: Real-Time VLAs via Future-State-Aware Asynchronous Inference | ✓ Link | 97.3 | VLASH (π0.5, async delay 2, LIBERO-Goal) | 2025-11-30 |
| Learning a Unified Latent Action Space from Videos with Action-centric Cycle Consistency | | 97.3 | CycleMimic (Full) | 2026-06-03 |
| Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbone | ✓ Link | 97.2 | Dream-VLA | 2025-12-27 |
| Causal World Modeling for Robot Control | ✓ Link | 97.2 | LingBot-VA | 2026-01-29 |
| VLA-JEPA: Enhancing Vision-Language-Action Model with Latent World Model | ✓ Link | 97.2 | VLA-JEPA | 2026-02-10 |
| Dual Latent Memory in Vision-Language-Action Models for Robotic Manipulation | ✓ Link | 97.2 | LaMem-VLA | 2026-07-08 |
| GEM: Generative Supervision Helps Embodied Intelligence | ✓ Link | 97.1 | GEM-VLA | 2026-05-27 |
| VisualThink-VLA: Visual Intermediate Reasoning for Effective and Low-Latency Vision-Language-Action Policies | ✓ Link | 97.05 | VisualThink-VLA | 2026-05-28 |
| Fast-WAM: Do World Action Models Need Test-time Future Imagination? | ✓ Link | 97.0 | Fast-WAM | 2026-03-17 |
| DRIFT: A Residual Flow Adapter for Decoding Continuous Outputs in Vision-Language Models | ✓ Link | 97.0 | DRIFT (Qwen3-VL-2B) | 2026-06-04 |
| RhinoVLA Technical Report | ✓ Link | 97.0 | RhinoVLA + VR Pretrain | 2026-06-05 |
| Discrete Diffusion VLA: Bringing Discrete Diffusion to Action Decoding in Vision-Language-Action Policies | ✓ Link | 96.8 | Discrete Diffusion VLA | 2025-08-27 |
| Recurrent-Depth VLA: Implicit Test-Time Compute Scaling of Vision-Language-Action Models via Latent Iterative Reasoning | ✓ Link | 96.8 | RD-VLA (Adaptive) | 2026-02-08 |
| Enfold: Folding World Model Imagination into Predictive Representations for Ultra-Efficient Embodied Control | ✓ Link | 96.8 | Enfold | 2026-08-06 |
| GeoVLA: Empowering 3D Representations in Vision-Language-Action Models | | 96.6 | GeoVLA | 2025-08-12 |
| CogVLA: Cognition-Aligned Vision-Language-Action Model via Instruction-Driven Routing & Sparsification | ✓ Link | 96.6 | CogVLA | 2025-08-28 |
| VLANeXt: Recipes for Building Strong VLA Models | ✓ Link | 96.6 | VLANeXt | 2026-02-20 |
| InstructVLA: Vision-Language-Action Instruction Tuning from Understanding to Manipulation | ✓ Link | 96.5 | InstructVLA-1.5B | 2025-07-23 |
| MemoryVLA: Perceptual-Cognitive Memory in Vision-Language-Action Models for Robotic Manipulation | | 96.4 | MemoryVLA | 2025-08-26 |
| DeepThinkVLA: Enhancing Reasoning Capability of Vision-Language-Action Models | ✓ Link | 96.4 | DeepThinkVLA (pi0-FAST) | 2025-10-31 |
| RynnVLA-002: A Unified Vision-Language-Action and World Model | ✓ Link | 96.4 | RynnVLA-002 | 2025-11-21 |
| Flash-WAM: Modality-Aware Distillation for World Action Models | ✓ Link | 96.4 | Flash-WAM (1v/2a) | 2026-06-03 |
| APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies | ✓ Link | 96.4 | APT | 2026-06-10 |
| VLA-0: Building State-of-the-Art VLAs with Zero Modification | ✓ Link | 96.2 | VLA-0 | 2025-10-15 |
| AffordanceVLA: A Vision-Language-Action Model Empowering Action Generation through Affordance-Aware Understanding | ✓ Link | 96.2 | AffordanceVLA (full) | 2026-06-04 |
| FLOWER: Democratizing Generalist Robot Policies with Efficient Vision-Language-Action Flow Policies | ✓ Link | 96.1 | FLOWER | 2025-09-05 |
| DepthVLA: Enhancing Vision-Language-Action Models with Depth-Aware Spatial Reasoning | | 95.8 | DepthVLA | 2025-10-15 |
| MemBodied: Recurrent Associative Memory for Vision-Language-Action Models | ✓ Link | 95.8 | MemBodied | 2026-09-23 |
| DexVLA: Vision-Language Model with Plug-In Diffusion Expert for General Robot Control | ✓ Link | 95.6 | DexVLA | 2025-02-09 |
| UniVLA: Learning to Act Anywhere with Task-centric Latent Actions | ✓ Link | 95.6 | UniVLA (Task-centric Latent Actions) | 2025-05-09 |
| Knowledge Insulating Vision-Language-Action Models: Train Fast, Run Fast, Generalize Better | | 95.6 | Knowledge-Insulated VLA (from generalist model) | 2025-05-29 |
| StructRL: Online Structured Reinforcement Learning for Long-Horizon Vision-Language-Action Tasks | ✓ Link | 95.6 | StructRL (GR00T-N1.5) | 2026-09-28 |
| BitVLA: 1-bit Vision-Language-Action Models for Robotics Manipulation | ✓ Link | 95.4 | BitVLA | 2025-06-09 |
| F1: A Vision-Language-Action Model Bridging Understanding and Generation to Actions | ✓ Link | 95.4 | F1 (pretrained) (LIBERO-Goal) | 2025-09-08 |
| VLA-RFT: Vision-Language-Action Reinforcement Fine-tuning with Verified Rewards in World Simulators | ✓ Link | 95.4 | VLA-RFT | 2025-10-01 |
| NORA-1.5: A Vision-Language-Action Model Trained using World Model- and Action-based Preference Rewards | ✓ Link | 95.4 | NORA-1.5 (DPO) (LIBERO-Goal) | 2025-11-18 |
| StaMo: Unsupervised Learning of Generalizable Robot Motion from Compact State Representation | ✓ Link | 95.0 | OpenVLA-OFT + StaMo state | 2025-10-06 |
| Accelerating Vision-Language-Action Model Integrated with Action Chunking via Parallel Decoding | | 94.9 | PD-VLA | 2025-03-04 |
| MM-ACT: Learn from Multimodal Parallel Generation to Act | ✓ Link | 94.8 | MM-ACT (Vanilla) | 2025-11-30 |
| Action-Sketcher: From Reasoning to Action via Visual Sketches for Robotic Manipulation | ✓ Link | 94.8 | Action-Sketcher (extra data) | 2026-01-04 |
| Chain of World: World Model Thinking in Latent Motion | ✓ Link | 94.6 | Chain of World | 2026-03-03 |
| Mantis: A Versatile Vision-Language-Action Model with Disentangled Visual Foresight | ✓ Link | 94.4 | Mantis | 2025-11-20 |
| InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policy | ✓ Link | 93.8 | InternVLA-M1 (LIBERO-Goal) | 2025-10-15 |
| Unified Vision-Language-Action Model | ✓ Link | 93.6 | UniVLA (LIBERO-Goal) | 2025-06-24 |
| Beyond Dense Futures: World Models as Structured Planners for Robotic Manipulation | ✓ Link | 93.4 | StructVLA | 2026-03-13 |
| Dita: Scaling Diffusion Transformer for Generalist Vision-Language-Action Policy | ✓ Link | 93.2 | Dita w/ wrist (LIBERO-Goal) | 2025-03-25 |
| WorldDiT: A Unified Diffusion Architecture for World and Action Modeling | | 92.8 | WorldDiT | 2026-07-27 |
| SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics | ✓ Link | 92 | SmolVLA (0.45B) | 2025-06-02 |
| DynBridge: Bridging Imagination and Control through Interaction Dynamics for Robot Manipulation | ✓ Link | 92.0 | DynBridge | 2026-06-03 |
| villa-X: Enhancing Latent Action Modeling in Vision-Language-Action Models | ✓ Link | 91.5 | villa-X (LIBERO-Goal) | 2025-07-31 |
| Reinforcing Action Policies by Prophesying | ✓ Link | 91.2 | VLA-Adapter + FA-GRPO & FlowScale | 2025-11-25 |
| SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics | ✓ Link | 91 | SmolVLA (2.25B) | 2025-06-02 |
| 4D-VLA: Spatiotemporal Vision-Language-Action Pretraining with Cross-Scene Calibration | | 90.9 | 4D-VLA | 2025-06-27 |
| CorridorVLA: Explicit Spatial Constraints for Generative Action Heads via Sparse Anchors | | 90.8 | SmolVLA-Corr | 2026-04-23 |
| CLAP: Direct VLM-to-VLA Adaptation via Language-Action Grounding | ✓ Link | 90.8 | CLAP Qwen3.5 2B | 2026-07-09 |
| mimic-video: Video-Action Models for Generalizable Robot Control Beyond VLAs | ✓ Link | 90.6 | mimic-video | 2025-12-17 |
| DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledge | ✓ Link | 89.5 | DreamVLA (LIBERO-Goal) | 2025-07-06 |
| NORA: A Small Open-Sourced Generalist Vision Language Action Model for Embodied Tasks | ✓ Link | 89.4 | NORA-Long-fine-tuned (LIBERO-Goal) | 2025-04-28 |
| G^3VLA: Geometric Inductive Bias for Vision-Language-Action Models | | 88.4 | G^3VLA (GT) | 2026-10-14 |
| CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Models | | 87.6 | CoT-VLA-7B | 2025-03-27 |
| MolmoAct: Action Reasoning Models that can Reason in Space | ✓ Link | 87.6 | MolmoAct-7B-D (extra data) | 2025-08-11 |
| ThinkAct: Vision-Language-Action Reasoning via Reinforced Visual Latent Planning | | 87.1 | ThinkAct | 2025-07-22 |
| Diffusion Transformer Policy | ✓ Link | 85.4 | DiT Policy (LIBERO-Goal) | 2024-10-21 |
| Dita: Scaling Diffusion Transformer for Generalist Vision-Language-Action Policy | ✓ Link | 85.4 | Dita (third-person RGB) (LIBERO-Goal) | 2025-03-25 |
| WorldVLA: Towards Autoregressive Action World Model | ✓ Link | 83.4 | WorldVLA (512x512) | 2025-06-26 |
| VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learning | ✓ Link | 82.2 | VLA-RL (LIBERO-Goal) | 2025-05-24 |
| OTTER: A Vision-Language-Action Model with Text-Aware Visual Feature Extraction | ✓ Link | 79.0 | OTTER | 2025-03-05 |
| Visual Spatial Tuning | ✓ Link | 76.0 | VST-tuned Qwen2.5-VL-3B (10K) | 2025-11-07 |
| VQ-VLA: Improving Vision-Language-Action Models via Scaling Vector-Quantized Action Tokenizers | ✓ Link | 75.2 | VQ-VLA (Conv Residual VQ-VAE, all-LIBERO) | 2025-07-01 |
| TraceVLA: Visual Trace Prompting Enhances Spatial-Temporal Awareness for Generalist Robotic Policies | | 75.1 | TraceVLA finetuned | 2024-12-13 |