Runway 发布 Praxis-1:机械臂任务速度比 GPT-6 Astra 快 15.6 倍
Runway 不做视频了?他们发的 Praxis-1 是机器人操控模型,驱动机械臂比用 GPT-6 Astra 快 15.6 倍、单任务便宜 51 倍。做具身智能或产线自动化的开发者可以关注专用模型和通用 LLM 这条路线之争。
Runway 宣布推出 Praxis-1,一款面向机器人操控的模型。在同一测试站和任务集上,Praxis-1 驱动机械臂完成任务的耗时只有 GPT-6 Astra 的 1/15.6,单任务成本降低 51 倍。
从公布的数据看,Praxis-1 走的是专用操控模型路线,而非把通用大语言模型接到机械臂上。通用 LLM 在规划与对话上强,但生成的动作序列延迟高、推理成本大,这类专用模型通常通过针对操控场景的训练压缩推理开销,适合高频重复的抓取与装配类任务。
这组对比数字也暗示机器人赛道正在分化:一边是 GPT-6 Astra 这样的通用模型驱动硬件,一边是 Praxis-1 这类专攻速度与单位成本的垂直方案。对做具身智能和自动化产线的团队来说,单任务成本 51 倍的差距直接决定商业上能否落地。
Praxis-1 dramatically outperforms frontier LLMs on speed and cost. On the same station and tasks, Praxis-1 finished 15.6x faster than GPT-6 Astra driving the arm, at 51x lower cost per completed task.
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Praxis-1, Runway, GPT-6 Astra, embodied AI.
Source attribution: curated from Runway ML (@runwayml). Cross-source corroboration: 11 independent sources have covered this development (including Artificial Analysis, IT之家, marktechpost), confirming industry-wide signal strength.