VLA MODELS · LANDSCAPE

VLA models differ most in action representation, embodiment scope, data and deployment assumptions.

There is no single VLA architecture. Current Vision-Language-Action models range from open token-based policies to compact continuous-action models, on-device systems and large proprietary robot foundation models.

CITATION RECORD
MATURITYVLA MODEL LANDSCAPE · SOURCE-BACKED REFERENCE
LAST VERIFIED27 Sep 2026
SOURCE VERSIONPrimary model pages and model cards current to Sep 2026
EVIDENCE BOUNDARYRepresentative, not exhaustive. Model capabilities are reported from primary sources and are not normalized benchmark claims.

HOW TO COMPARE

Model size alone is a poor way to compare VLA systems.

A useful VLA comparison asks what observations the policy receives, whether proprioception is included, how actions are represented, whether the model predicts single steps or chunks, which embodiments it was trained on, how it is adapted to a new robot, and what compute assumptions exist at inference time.

REPRESENTATIVE MODELS

A practical map of the current VLA landscape.

OPEN · 7B

OpenVLA

Open-source VLA pretrained on 970k Open X-Embodiment episodes. The original model predicts tokenized actions that are decoded into continuous robot actions.

OPEN · 450M

SmolVLA

Compact LeRobot VLA using SmolVLM2 plus a continuous action expert. Designed for consumer hardware, SO-100/SO-101-class robots and asynchronous inference.

PROPRIETARY

Gemini Robotics 2

Google DeepMind VLA family with image and text inputs and robot actions as outputs, designed for broad robot capability and multi-embodiment use.

ON-DEVICE

Gemini Robotics On-Device 2

Local VLA accepting text, images and numerical proprioception, returning numerical robot actions and targeting efficient on-device manipulation.

HUMANOID

GR00T N1.x

NVIDIA foundation-model family for humanoid robots. N1.5 combines a VLM with a diffusion-transformer action path and flow-matching policy learning.

FLOW MODEL

π0

Physical Intelligence VLA that combines vision, language and robot state with a continuous action expert trained using conditional flow matching.

OPEN-WORLD

π0.5

Extension of π0 toward mobile manipulation and unseen homes, combining heterogeneous robot data, semantic supervision and web-derived multimodal data.

OPENVLA

OpenVLA made the VLA stack reproducible at meaningful scale.

OpenVLA is a 7B model built from a fused visual encoder, projector and Llama-family language backbone. Its original action interface discretizes robot actions into tokens, allowing the language-model backbone to predict them autoregressively. The project later added faster adaptation and inference techniques, including FAST-style action tokenization and continuous-action optimized fine-tuning.

SMOLVLA

SmolVLA shows that VLA research does not require a giant model.

Hugging Face reports 450M parameters for SmolVLA. It uses a VLM backbone plus an action expert, takes multiple RGB views, sensorimotor state and a language instruction, and generates a continuous sequence of actions. Its open LeRobot integration and SO-101 evaluation make it especially relevant for accessible real-robot experimentation.

GEMINI ROBOTICS

Google DeepMind pushes VLA toward broad multi-embodiment deployment.

Gemini Robotics 2 is presented as a VLA with text and image inputs and actions as outputs. The On-Device 2 model card makes the embodiment interface particularly explicit: text, images and numerical robot proprioception enter the model, and numerical robot actions come out.

GR00T

GR00T couples language grounding, robot state and continuous action generation.

NVIDIA GR00T N1.5 uses a vision-language backbone to encode text and visual observations while a diffusion-transformer path processes robot state and noised actions. NVIDIA also adds future-latent representation learning, illustrating how policy learning and predictive world-model objectives can increasingly coexist in one robot foundation model.

π0 / π0.5

Physical Intelligence uses continuous flow-based action generation.

π0 conditions an action expert on images, language and robot state and trains it with conditional flow matching. π0.5 adds heterogeneous sources and hierarchical semantic subtasks to support longer mobile-manipulation tasks and broader generalization.

COMPARISON BOUNDARY

Do not treat benchmark numbers from different VLA papers as one leaderboard.

VLA papers frequently use different robots, datasets, camera layouts, control rates, task suites, fine-tuning budgets and success definitions. OpenRGD therefore treats model-page metrics as source-specific evidence unless a benchmark protocol directly normalizes the comparison.

PRIMARY SOURCES