Meta's MuseSpark 1.2 Gains Embodied Skills: Robot Navigation, Table Cleaning, and Visual Coding
量子位 · wechat · 2026-08-21
Meta has showcased the multimodal and embodied capabilities of MuseSpark 1.2, demonstrating its ability to control robots and generate interfaces from visuals.
Architecture and Demos:
The system features a two-layer architecture. An upper-layer MuseSpark variant handles high-level planning and task decomposition, while a lower-layer VLA (Vision-Language-Action) model executes physical actions. Demos include a robot searching a room for a duck and a dual-arm robot organizing a desk. The model iteratively refines actions based on real-time visual feedback.
Key Capabilities:
- Visual Coding: Can generate websites and games directly from images or videos, ranking first on the DesignArena "video-to-website" leaderboard.
- Embodied Agent: Achieved 54.0% on the ZeroBench visual reasoning test, closely trailing GPT-5.6 Sol (54.6%).
- WildArtifactBench: A new internal benchmark covering 10 complex tasks (e.g., bird call recognition, heart ultrasound analysis, 3D mesh generation) to evaluate cross-modal processing and structured output.
Availability: Currently used internally for media generation and data annotation. Accessible via Meta Model API and MuseCode, with an open-source release confirmed.
More from Embodied
- Tesla officially unveils new flying Roadster, prompting Grok barrel-roll memes — EXM7777 · 2026-10-16
- Tesla says new Roadster's aero package is driven by Optimus actuators, nearing F1 downforce — yunta_tsai · 2026-10-16
- Tesla Roadster event end sizzle reel surfaces online — banteg · 2026-10-16
- Up-close photos reveal the Roadster's cold-gas thruster tank and nozzles — PTrubey · 2026-10-16
- Real-time collaborative mapping across phones and robots on the Auki network — broodsugar · 2026-10-16
- Tesla Roadster does 0-60 mph in 0.81 seconds — not a typo — Polymarket · 2026-10-16