NVIDIA's VFlash: fast visual planning VLM for physical AI, 5x faster on Jetson Thor
ZhidingYu · x · 2026-11-06
NVIDIA's team (Yu Zhiding et al.) releases VFlash, a fast visual-planning VLM for physical AI, aiming to make visual planning fast, accurate, and directly actionable — a step toward System-1 VLMs.
Built on NVIDIA's TensorRT-Edge-LLM and TensorRT stacks, VFlash deploys across Jetson Thor, DGX Spark, and H100: on Jetson Thor and DGX Spark it delivers 5x faster inference on MolmoSpace and over 3x on other workloads, covering robot manipulation and autonomous driving.
Technical report, project page, and code are all available.
More from Embodied
- GPT-6 Astra tops a 16-benchmark robot index, yet went 0/11 on real plug insertions — AIFlow_ML · 2026-11-06
- Giving Claude a cutting-edge robot body just so it can spoon-feed me — repligate · 2026-11-06
- moonlake previews research on scaling robot learning with simulation in the RL era — du_yilun · 2026-11-06
- Nearly 100 testers praise Sphenomenon MR demo that turns a whole house into voxels — Scobleizer · 2026-11-06
- Spectral Skills lands on real robots: composing motion skills via energy functions — daibond_alpha · 2026-11-06
- Developer gives Claude Desktop a body and sensors via MCP and Raspberry Pi — jfrankie · 2026-11-06