Skip to content
Back to search

VLA-Arena

VLA-Arena is an open-source benchmark for systematic evaluation of Vision-Language-Action (VLA) models.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/pku-alignment--vla-arena/badge.svg)](/listings/pku-alignment--vla-arena)

HTML

<a href="/listings/pku-alignment--vla-arena"><img src="/api/listings/pku-alignment--vla-arena/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

> **Prerequisite**: install uv: https://docs.astral.sh/uv/

Step 1 — Clone

git clone https://github.com/PKU-Alignment/VLA-Arena.git
cd VLA-Arena

Step 2 — Run (Evaluate or Train)

You can directly evaluate using our official finetuned models, or train your own. *(The first `uv run` may take a while as it automatically creates the isolated environment and installs dependencies).*

**To Evaluate:**

uv run --project envs/openvla \
  vla-arena eval --model openvla --config vla_arena/configs/evaluation/openvla.yaml

**To Train:**

uv run --project envs/openvla \
  vla-arena train --model openvla --config vla_arena/configs/train/openvla.yaml

---

⚙️ Configuration

Before running the commands above, edit the YAML configs for your model setup. Example (OpenVLA):

* **Training Config** (`vla_arena/configs/train/openvla.yaml`): Set `vla_path`, `data_root_dir`, and `dataset_name`. * **Evaluation Config** (`vla_arena/configs/evaluation/openvla.yaml`): Set `pretrained_checkpoint`, `task_suite_name`, and `task_level`.

Other models follow the same pattern: use the matching `vla_arena/configs/train/<model>.yaml`, `vla_arena/configs/evaluation/<model>.yaml`, and `envs/<model>`.

> 💡 For data collection and dataset conversion, see `docs/data_collection.md`.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 375 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. <h1 align="center">🤖 VLA-Arena: An Open-Source Framework for Benchmarking Vision-Language-Action Models</h1>
  2. <p align="center">
  3. <a href="https://arxiv.org/abs/2512.22539"><img src="https://img.shields.io/badge/arXiv-2512.22539-B31B1B?style=for-the-badge&link=https%3A%2F%2Farxiv.org%2Fabs%2F2512.22539" alt="arXiv"></a>
  4. <a href="LICENSE"><img src="https://img.shields.io/badge/license-%20Apache%202.0-green?style=for-the-badge" alt="License"></a>
  5. <a href="https://www.python.org/downloads/"><img src="https://img.shields.io/badge/python-3.11-blue?style=for-the-badge" alt="Python"></a>
  6. <a href="https://vla-arena.github.io/#leaderboard"><img src="https://img.shields.io/badge/leaderboard-available-purple?style=for-the-badge" alt="Leaderboard"></a>
  7. <a href="https://vla-arena.github.io/#taskstore"><img src="https://img.shields.io/badge/task%20store-170+%20tasks-orange?style=for-the-badge" alt="Task Store"></a>
  8. <a href="https://huggingface.co/vla-arena"><img src="https://img.shields.io/badge/🤗%20models%20%26%20datasets-available-yellow?style=for-the-badge" alt="Models & Datasets"></a>
  9. <a href="docs/"><img src="https://img.shields.io/badge/docs-available-green?style=for-the-badge" alt="Docs"></a>
  10. <a href="https://github.com/vla-arena/vla-arena-community"><img src="https://img.shields.io/badge/Community-💬%20WeChat%20%26%20𝕏-cyan?style=for-the-badge" alt="Community"></a>
  11. </p>
  12. <div align="center">
  13. <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/logo.jpeg" width="75%"/>
  14. </div>
  15. VLA-Arena is an open-source benchmark for systematic evaluation of Vision-Language-Action (VLA) models. VLA-Arena provides a full toolchain covering *scenes modeling*, *demonstrations collection*, *models training* and *evaluation*. It features 170 tasks across 11 specialized suites, hierarchical difficulty levels (L0-L2), and comprehensive metrics for safety, generalization, and efficiency assessment.
  16. VLA-Arena focuses on four key domains:
  17. - **Safety**: Operate reliably and safely in the physical world.
  18. - **Distractors**: Maintain stable performance when facing environmental unpredictability.
  19. - **Extrapolation**: Generalize learned knowledge to novel situations.
  20. - **Long Horizon**: Combine long sequences of actions to achieve a complex goal.
  21. ## 📰 News
  22. - **[2025.12.27]** 📄 Our [paper](https://arxiv.org/abs/2512.22539) is now available!
  23. - **[2025.09.29]** 🚀 VLA-Arena is officially released!
  24. ## 🔥 Highlights
  25. - **🚀 End-to-End & Out-of-the-Box**: We provide a complete and unified toolchain covering everything from scene modeling and behavior collection to model training and evaluation. Paired with comprehensive docs and tutorials, you can get started in minutes.
  26. - **🔌 Plug-and-Play Evaluation**: Seamlessly integrate and benchmark your own VLA models. Our framework is designed with a unified API, making the evaluation of new architectures straightforward with minimal code changes.
  27. - **🛠️ Effortless Task Customization**: Leverage the Constrained Behavior Domain Definition Language (CBDDL) to rapidly define entirely new tasks and safety constraints. Its declarative nature allows you to achieve comprehensive scenario coverage with minimal effort.
  28. - **📊 Systematic Difficulty Scaling**: Systematically assess model capabilities across three distinct difficulty levels (L0→L1→L2). Isolate specific skills and pinpoint failure points, from basic object manipulation to complex, long-horizon tasks.
  29. ## 📚 Table of Contents
  30. - [Quick Start](#quick-start)
  31. - [Task Suites Overview](#task-suites-overview)
  32. - [Installation](#installation)
  33. - [Documentation](#documentation)
  34. - [Leaderboard](#leaderboard)
  35. - [Contributing](#contributing)
  36. - [License](#license)
  37. ## Quick Start
  38. > **Prerequisite**: install uv: https://docs.astral.sh/uv/
  39. ### Step 1 — Clone
  40. ```bash
  41. git clone https://github.com/PKU-Alignment/VLA-Arena.git
  42. cd VLA-Arena
  43. ```
  44. ### Step 2 — Run (Evaluate or Train)
  45. You can directly evaluate using our official finetuned models, or train your own. *(The first `uv run` may take a while as it automatically creates the isolated environment and installs dependencies).*
  46. **To Evaluate:**
  47. ```bash
  48. uv run --project envs/openvla \
  49. vla-arena eval --model openvla --config vla_arena/configs/evaluation/openvla.yaml
  50. ```
  51. **To Train:**
  52. ```bash
  53. uv run --project envs/openvla \
  54. vla-arena train --model openvla --config vla_arena/configs/train/openvla.yaml
  55. ```
  56. ---
  57. ### ⚙️ Configuration
  58. Before running the commands above, edit the YAML configs for your model setup. Example (OpenVLA):
  59. * **Training Config** (`vla_arena/configs/train/openvla.yaml`): Set `vla_path`, `data_root_dir`, and `dataset_name`.
  60. * **Evaluation Config** (`vla_arena/configs/evaluation/openvla.yaml`): Set `pretrained_checkpoint`, `task_suite_name`, and `task_level`.
  61. Other models follow the same pattern: use the matching `vla_arena/configs/train/<model>.yaml`, `vla_arena/configs/evaluation/<model>.yaml`, and `envs/<model>`.
  62. > 💡 For data collection and dataset conversion, see `docs/data_collection.md`.
  63. ## Task Suites Overview
  64. VLA-Arena provides 11 specialized task suites with 170 tasks total, organized into four domains:
  65. ### 🛡️ Safety (5 suites, 75 tasks)
  66. | Suite | Description | L0 | L1 | L2 | Total |
  67. |-------|------------|----|----|----|-------|
  68. | `safety_static_obstacles` | Static collision avoidance | 5 | 5 | 5 | 15 |
  69. | `safety_cautious_grasp` | Safe grasping strategies | 5 | 5 | 5 | 15 |
  70. | `safety_hazard_avoidance` | Hazard area avoidance | 5 | 5 | 5 | 15 |
  71. | `safety_state_preservation` | Object state preservation | 5 | 5 | 5 | 15 |
  72. | `safety_dynamic_obstacles` | Dynamic collision avoidance | 5 | 5 | 5 | 15 |
  73. ### 🔄 Distractor (2 suites, 30 tasks)
  74. | Suite | Description | L0 | L1 | L2 | Total |
  75. |-------|------------|----|----|----|-------|
  76. | `distractor_static_distractors` | Cluttered scene manipulation | 5 | 5 | 5 | 15 |
  77. | `distractor_dynamic_distractors` | Dynamic scene manipulation | 5 | 5 | 5 | 15 |
  78. ### 🎯 Extrapolation (3 suites, 45 tasks)
  79. | Suite | Description | L0 | L1 | L2 | Total |
  80. |-------|------------|----|----|----|-------|
  81. | `preposition_combinations` | Spatial relationship understanding | 5 | 5 | 5 | 15 |
  82. | `task_workflows` | Multi-step task planning | 5 | 5 | 5 | 15 |
  83. | `unseen_objects` | Unseen object recognition | 5 | 5 | 5 | 15 |
  84. ### 📈 Long Horizon (1 suite, 20 tasks)
  85. | Suite | Description | L0 | L1 | L2 | Total |
  86. |-------|------------|----|----|----|-------|
  87. | `long_horizon` | Long-horizon task planning | 10 | 5 | 5 | 20 |
  88. **Difficulty Levels:**
  89. - **L0**: Basic tasks with clear objectives
  90. - **L1**: Intermediate tasks with increased complexity
  91. - **L2**: Advanced tasks with challenging scenarios
  92. ### 🛡️ Safety Suites Visualization
  93. | Suite Name | L0 | L1 | L2 |
  94. |------------|----|----|----|
  95. | **Static Obstacles** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/static_obstacles_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/static_obstacles_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/static_obstacles_2.png" width="175" height="175"> |
  96. | **Cautious Grasp** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/safe_pick_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/safe_pick_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/safe_pick_2.png" width="175" height="175"> |
  97. | **Hazard Avoidance** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/dangerous_zones_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/dangerous_zones_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/dangerous_zones_2.png" width="175" height="175"> |
  98. | **State Preservation** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/task_object_state_maintenance_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/task_object_state_maintenance_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/task_object_state_maintenance_2.png" width="175" height="175"> |
  99. | **Dynamic Obstacles** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/dynamic_obstacle_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/dynamic_obstacle_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/dynamic_obstacle_2.png" width="175" height="175"> |
  100. ### 🔄 Distractor Suites Visualization
  101. | Suite Name | L0 | L1 | L2 |
  102. |------------|----|----|----|
  103. | **Static Distractors** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/robustness_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/robustness_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/robustness_2.png" width="175" height="175"> |
  104. | **Dynamic Distractors** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/moving_obstacles_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/moving_obstacles_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/moving_obstacles_2.png" width="175" height="175"> |
  105. ### 🎯 Extrapolation Suites Visualization
  106. | Suite Name | L0 | L1 | L2 |
  107. |------------|----|----|----|
  108. | **Preposition Combinations** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/preposition_generalization_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/preposition_generalization_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/preposition_generalization_2.png" width="175" height="175"> |
  109. | **Task Workflows** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/workflow_generalization_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/workflow_generalization_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/workflow_generalization_2.png" width="175" height="175"> |
  110. | **Unseen Objects** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/unseen_object_generalization_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/unseen_object_generalization_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/unseen_object_generalization_2.png" width="175" height="175"> |
  111. ### 📈 Long Horizon Suite Visualization
  112. | Suite Name | L0 | L1 | L2 |
  113. |------------|----|----|----|
  114. | **Long Horizon** | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/long_horizon_0.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/long_horizon_1.png" width="175" height="175"> | <img src="https://raw.githubusercontent.com/PKU-Alignment/VLA-Arena/main/image/long_horizon_2.png" width="175" height="175"> |
  115. ## Installation
  116. ### System Requirements
  117. - **OS**: Ubuntu 20.04+ or macOS 12+
  118. - **Python**: 3.11.x (`==3.11.*`)
  119. - **CUDA**: 11.8+ (for GPU acceleration)
  120. ### Install from Source (Recommended)
  121. ```bash
  122. # Clone repository
  123. git clone https://github.com/PKU-Alignment/VLA-Arena.git
  124. cd VLA-Arena
  125. # Install uv: https://docs.astral.sh/uv/
  126. # (Optional) Pre-install base environment (otherwise the first `uv run` will do it)
  127. uv sync --project envs/base
  128. # (Optional) Download / update task suites and assets from the Hub (~850 MB)
  129. uv run --project envs/base vla-arena.download-tasks install-all --repo vla-arena/tasks
  130. ```
  131. > **Note**: If you cloned this repository, tasks and assets are already included. You can skip the download step unless you want to update from the Hub.
  132. ### Install from PyPI (Alternative)
  133. ```bash
  134. python3 -m pip install vla-arena
  135. # One-time: initialize local uv projects (`envs/*`) and copy default configs
  136. vla-arena.init-workspace --force
  137. # (Optional) Download task suites / assets (~850 MB)
  138. uv run --project envs/base vla-arena.download-tasks install-all --repo vla-arena/tasks
  139. # One-line train / eval (config auto-defaults; override via --config if needed)
  140. uv run --project envs/openvla vla-arena train --model openvla
  141. uv run --project envs/openvla vla-arena eval --model openvla
  142. ```
  143. For source checkout users, the existing `envs/<model_name>` workflow remains unchanged.
  144. ## Documentation
  145. VLA-Arena provides comprehensive documentation for all aspects of the framework. Choose the guide that best fits your needs:
  146. ### 📖 Core Guides
  147. #### 🏗️ [Scene Construction Guide](docs/scene_construction.md) | [中文版](docs/scene_construction_zh.md)
  148. Build custom task scenarios using CBDDL (Constrained Behavior Domain Definition Language).
  149. - CBDDL file structure and syntax
  150. - Region, fixture, and object definitions
  151. - Moving objects with various motion types (linear, circular, waypoint, parabolic)
  152. - Initial and goal state specifications
  153. - Cost constraints and safety predicates
  154. - Image effect settings
  155. - Asset management and registration
  156. - Scene visualization tools
  157. #### 📊 [Data Collection Guide](docs/data_collection.md) | [中文版](docs/data_collection_zh.md)
  158. Collect demonstrations in custom scenes and convert data formats.
  159. - Interactive simulation environment with keyboard controls
  160. - Demonstration data collection workflow
  161. - Data format conversion (HDF5 to training dataset)
  162. - Dataset regeneration (filtering noops and optimizing trajectories)
  163. - Convert dataset to RLDS format (for X-embodiment frameworks)
  164. - Convert RLDS dataset to LeRobot format (for Hugging Face LeRobot)
  165. #### 🔧 [Model Fine-tuning and Evaluation Guide](docs/finetuning_and_evaluation.md) | [中文版](docs/finetuning_and_evaluation_zh.md)
  166. Fine-tune and evaluate VLA models using VLA-Arena generated datasets.
  167. - Unified uv-only workflow for all supported models
  168. - Per-model isolated environments (`envs/openvla`, `envs/openvla_oft`, `envs/univla`, `envs/smolvla`, `envs/openpi`)
  169. - Training configuration and hyperparameter settings
  170. - Evaluation scripts and metrics
  171. - Policy server setup for inference (OpenPi)
  172. ### 🔜 Quick Reference
  173. #### Common Commands
  174. - **Train**: `uv run --project envs/<model_name> vla-arena train --model <model_cli_name>` (optional override: `--config ...`)
  175. - **Eval**: `uv run --project envs/<model_name> vla-arena eval --model <model_cli_name>` (optional override: `--config ...`)
  176. - See the [Model Fine-tuning and Evaluation Guide](docs/finetuning_and_evaluation.md).
  177. #### Documentation Index
  178. - **English**: [`README_EN.md`](docs/README_EN.md) - Complete English documentation index
  179. - **中文**: [`README_ZH.md`](docs/README_ZH.md) - 完整中文文档索引
  180. ### 📦 Download Task Suites
  181. #### Method 1: Using CLI Tool (Recommended)
  182. After installation, you can use the following commands to view and download task suites:
  183. ```bash
  184. # View installed tasks
  185. uv run --project envs/base vla-arena.download-tasks installed
  186. # List available task suites
  187. uv run --project envs/base vla-arena.download-tasks list --repo vla-arena/tasks
  188. # Install a single task suite
  189. uv run --project envs/base vla-arena.download-tasks install distractor_dynamic_distractors --repo vla-arena/tasks
  190. # Install multiple task suites at once
  191. uv run --project envs/base vla-arena.download-tasks install safety_hazard_avoidance safety_state_preservation --repo vla-arena/tasks
  192. # Install all task suites (recommended)
  193. uv run --project envs/base vla-arena.download-tasks install-all --repo vla-arena/tasks
  194. ```
  195. #### Method 2: Using Python Script
  196. ```bash
  197. # View installed tasks
  198. uv run --project envs/base python -m scripts.download_tasks installed
  199. # Install all tasks
  200. uv run --project envs/base python -m scripts.download_tasks install-all --repo vla-arena/tasks
  201. ```
  202. ### 🔧 Custom Task Repository
  203. If you want to use your own task repository:
  204. ```bash
  205. # Use custom HuggingFace repository
  206. uv run --project envs/base vla-arena.download-tasks install-all --repo your-username/your-task-repo
  207. ```
  208. ### 📝 Create and Share Custom Tasks
  209. You can create and share your own task suites:
  210. ```bash
  211. # Package a single task
  212. uv run --project envs/base vla-arena.manage-tasks pack path/to/task.bddl --output ./packages
  213. # Package all tasks
  214. uv run --project envs/base python scripts/package_all_suites.py --output ./packages
  215. # Upload to HuggingFace Hub
  216. uv run --project envs/base vla-arena.manage-tasks upload ./packages/my_task.vlap --repo your-username/your-repo
  217. ```
  218. ## Leaderboard
  219. ### Performance Evaluation of VLA Models on the VLA-Arena Benchmark
  220. We compare VLA models across four dimensions: **Safety**, **Distractor**, **Extrapolation**, and **Long Horizon**. Performance trends over three difficulty levels (L0–L2) are shown with a unified scale (0.0–1.0) for cross-model comparison. You can access detailed results and comparisons in our [leaderboard](https://vla-arena.github.io/#leaderboard).
  221. ---
  222. ## Sharing Research Results
  223. VLA-Arena provides a series of tools and interfaces to help you easily share your research results, enabling the community to understand and reproduce your work. This guide will introduce how to use these tools.
  224. ### 🤖 Sharing Model Results
  225. To share your model results with the community:
  226. 1. **Evaluate Your Model**: Evaluate your model on VLA-Arena tasks
  227. 2. **Submit Results**: Follow the [submission guidelines](https://github.com/vla-arena/vla-arena.github.io#contributing-your-model-results) in our leaderboard repository
  228. 3. **Create Pull Request**: Submit a pull request containing your model results
  229. ### 🎯 Sharing Task Designs
  230. Share your custom tasks through the following steps, enabling the community to reproduce your task configurations:
  231. 1. **Design Tasks**: Use CBDDL to [design your custom tasks](docs/scene_construction.md)
  232. 2. **Package Tasks**: Follow our guide to [package and submit your tasks](https://github.com/PKU-Alignment/VLA-Arena#-create-and-share-custom-tasks) to your custom HuggingFace repository
  233. 3. **Update Task Store**: Open a [Pull Request](https://github.com/vla-arena/vla-arena.github.io#contributing-your-tasks) to update your tasks in the VLA-Arena [task store](https://vla-arena.github.io/#taskstore)
  234. ## Contributing
  235. - **Report Issues**: Found a bug? [Open an issue](https://github.com/PKU-Alignment/VLA-Arena/issues)
  236. - **Improve Documentation**: Help us make the docs better
  237. - **Feature Requests**: Suggest new features or improvements
  238. ---
  239. ## Citing VLA-Arena
  240. If you find VLA-Arena useful, please cite it in your publications.
  241. ```bibtex
  242. @misc{zhang2025vlaarena,
  243. title={VLA-Arena: An Open-Source Framework for Benchmarking Vision-Language-Action Models},
  244. author={Borong Zhang and Jiahao Li and Jiachen Shen and Yishuai Cai and Yuhao Zhang and Yuanpei Chen and Juntao Dai and Jiaming Ji and Yaodong Yang},
  245. year={2025},
  246. eprint={2512.22539},
  247. archivePrefix={arXiv},
  248. primaryClass={cs.RO},
  249. url={https://arxiv.org/abs/2512.22539}
  250. }
  251. ```
  252. ---
  253. ## License
  254. This project is licensed under the Apache 2.0 license - see [LICENSE](LICENSE) for details.
  255. ## Acknowledgments
  256. - **RoboSuite**, **LIBERO**, and **VLABench** teams for the framework
  257. - **OpenVLA**, **UniVLA**, **Openpi**, and **lerobot** teams for pioneering VLA research
  258. - All contributors and the robotics community
  259. ---
  260. <p align="center">
  261. <b>VLA-Arena: An Open-Source Framework for Benchmarking Vision-Language-Action Models</b><br>
  262. Made with ❤️ by the VLA-Arena Team
  263. </p>

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.