aiqbench analysis
TurboVLA introduces a novel architecture that bypasses the computationally expensive LLM-centric pathway in favor of direct vision-language-to-action mapping, significantly reducing inference latency and memory footprint. By achieving real-time 32 Hz performance on consumer hardware with under 1 GB VRAM, this model offers a practical solution for deploying autonomous robotic agents where resource efficiency is critical. The work represents a meaningful shift toward lightweight yet high-performance models capable of bridging perception and action without relying on massive parameter counts.
TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with <1 GB VRAM
arXiv:2607.27205v1 Announce Type: cross Abstract: Vision-language-action (VLA) models commonly adopt an LLM-centric $V \to L \to A$ pathway, where visual observations are projected into the representation space of a large language model before being decoded into robot actions. Although effective, this design incurs substantial computation and memory overhead at every policy invocation. In this work, we introduce TurboVLA, a new VLA paradigm that reformulates the conventional $V \to L \to A$ pathway as a direct $V + L \to A$ mapping. Instead of using a large language model as the central interface between perception and action, TurboVLA independently encodes visual observations and language instructions, directly exchanges information between them through lightweight bidirectional vision-language interaction, and predicts continuous action chunks with a compact decoder. This simple design constructs task-conditioned representations directly from visual and linguistic features, significantly reducing the computational and memory costs of VLA inference. On LIBERO, TurboVLA achieves 97.7% average success with only 0.2B parameters, 31.2 ms inference latency, and 0.9 GB inference VRAM on a consumer-grade RTX 4090, matching or outperforming substantially larger VLA policies. These results establish TurboVLA as a simple and effective alternative to the prevailing LLM-centric VLA paradigm, offering a new perspective on how vision, language, and action can be connected for efficient robotic manipulation. Code is available at https://github.com/H-EmbodVis/TurboVLA.
Abstract:Vision-language-action (VLA) models commonly adopt an LLM-centric $V \to L \to A$ pathway, where visual observations are projected into the representation space of a large language model before being decoded into robot actions. Although effective, this design incurs substantial computation and memory overhead at every policy invocation. In this work, we introduce TurboVLA, a new VLA paradigm that reformulates the conventional $V \to L \to A$ pathway as a direct $V + L \to A$ mapping. Instead of using a large language model as the central interface between perception and action, TurboVLA independently encodes visual observations and language instructions, directly exchanges information between them through lightweight bidirectional vision-language interaction, and predicts continuous action chunks with a compact decoder. This simple design constructs task-conditioned representations directly from visual and linguistic features, significantly reducing the computational and memory costs of VLA inference. On LIBERO, TurboVLA achieves 97.7% average success with only 0.2B parameters, 31.2 ms inference latency, and 0.9 GB inference VRAM on a consumer-grade RTX 4090, matching or outperforming substantially larger VLA policies. These results establish TurboVLA as a simple and effective alternative to the prevailing LLM-centric VLA paradigm, offering a new perspective on how vision, language, and action can be connected for efficient robotic manipulation. Code is available at this https URL.
Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
arXiv:2607.27205 [cs.CV]
(or arXiv:2607.27205v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2607.27205 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 311 words
Related coverage
LeRobot v0.6.0: Imagine, Evaluate, Improve
Hugging Face · Jul 7, 2026
rMuscle: Robotic Muscle Memory for Efficient Vision-Language-Action Model Inference
arXiv · Sep 17, 2026
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · Oct 13, 2026
Building a Foundation Stack for General-Purpose Robots
IEEE Spectrum · Jul 13, 2026