Mistral AI 发布 Robostral Navigate：单摄像头导航模型

AI 摘要

Mistral AI 发布了 Robostral Navigate，一个 8B 参数的具身导航模型，仅使用单个 RGB 摄像头即可让机器人自主导航，在 R2R-CE 基准上达到 76.6% 的成功率，超越多传感器方法。该模型完全在模拟环境中训练，采用指向导航和强化学习，并计划扩展至统一具身智能体。

正文节选

Thinking Summary Robostral Navigate is an 8B model that enables robots to autonomously navigate complex environments using only a single RGB camera, achieving 76.6% success on unseen R2R-CE benchmarks—outperforming multi-sensor approaches while being more efficient. Built entirely in-house with simulated data and token-efficient techniques, it generalizes across robot types and adapts to real-world obstacles unseen during training. The model combines pointing-based navigation with reinforcement

发布时间：—

抓取时间：2026-08-02 00:22

来源机构：Mistral AI