독서리서치레이더투자 프레임워크
로그인 / 회원가입
← 읽기로 돌아가기
뉴스주제 · Praxis-1중대2 분 소요DisconfirmAI 편집부

Runway Releases Open-Source Praxis-1: A Video-Trained Robot Control Model Claiming Top Performance in Its Category

Runway releases Praxis-1, an open-source robot control model trained on video data, claiming leading results on Nvidia's RoboLab-120 benchmark, though unverified independently.

2026-11-14 17:45 GMT+8

AI company Runway released the open weights and source code for its robot control model, Praxis-1, on November 14, 2026, under the commercially permissive Apache 2.0 license. The model's core innovation lies in its "video-to-robot" transfer learning path: it first learns how objects move and interact from ordinary footage, then fine-tunes on videos of human actions and limited robot data. This allows it to predict environmental changes and calculate corresponding robot movements based on camera images, text instructions, and position data.

Runway calls Praxis-1 a "World Action Model," aiming to reduce reliance on expensive and time-consuming task-specific robot datasets. According to the company, adapting the model to a Franka robot arm required only about 277 hours of recordings from the public DROID dataset, while a dual-arm robot used between 1,400 and 2,600 hours of training material. Whether this approach actually needs less robot data overall than comparable models without general video pretraining remains an open question.

In terms of performance, Runway evaluated Praxis-1 itself on Nvidia's RoboLab-120 simulation benchmark, which tests sorting, grasping, or stacking across 120 tasks. The results show a 46.2 percent success rate, higher than FLUX 3 Action (42.9 percent), also built as a World Action Model, placing it at the top of that category. However, on the overall leaderboard, it ranks fifth behind two AI agents and two robot models with different architectures. It must be emphasized that all test results were run by Runway and have not been independently verified.

Beyond simulations, Runway conducted comparative experiments on a real Franka robot arm. When completing ten identical tasks (such as placing a banana on a plate or closing a drawer) with ten attempts each, Praxis-1 succeeded 88 times, while FLUX 3 Action managed 80. Additionally, in transfer tests on a mobile robot platform never seen before, after fine-tuning on about 95 hours of data, Praxis-1 hit an average success rate of 80 percent in simulated tests, significantly higher than Google's original RT-1 model's 52.4 percent.

로그인 / 회원가입
독서리서치레이더투자 프레임워크

Despite showing potential, Praxis-1 has clear limitations. According to Runway's research report, the model sometimes confuses similar-looking objects, struggles with vague instructions, and cannot reliably plan complex long-sequence actions autonomously, still requiring manual step decomposition. Currently, the standard model version requires a GPU with 80 GB of memory, while a compressed version is estimated to fit on an RTX 5090 (32 GB VRAM), though actual peak memory usage hasn't been measured. Developers can download the model from Hugging Face, with code hosted on GitHub.

출처

  • the-decoder.comThe Decoder↗
주제

Praxis-1

Runway于2026年11月正式开源其世界动作模型Praxis-1,标志着该公司将大规模视频预训练技术成功迁移至机器人控制领域。该模型基于270亿参数的混合专家扩散Transformer,通过预测场景变化来推导机器人动作,在Nvidia …

3 단계1 건업데이트됨 2026-11-18 09:40

현재까지의 과정

  1. 2026-09-01Runway首次宣布推出Praxis-1,作为其首个开放权重的世界动作模型,并向Noble Machines、Standard Bots等合作伙伴提供早期访问权限。runway.com ↗
  2. 2026-11-11Runway正式发布Praxis-1,以Apache 2.0许可证开源权重和代码,允许商业使用,并公布详细技术报告。runway.com ↗
  3. 2026-11-14媒体广泛报道Praxis-1在Nvidia RoboLab-120基准测试中取得46.2%成功率,领先于FLUX 3 Action等竞品,但指出结果未经独立验证。the-decoder.com ↗

아직 해결되지 않은 문제

  • 视频预训练是否真能如Runway所言,显著减少机器人策略训练所需的专用数据量?
  • 第三方机构能否在相同条件下复现Praxis-1在RoboLab-120上的46.2%成功率?
  • Praxis-1在处理复杂长序列任务和模糊指令时的局限性,是否会阻碍其在非结构化环境中的实际应用?
전체 주제 읽기 →
리서치 계속하기 →