Right-Click Anything: SAM 3.1, Muse Spark and Muse Image combined into one clever demo
nikhilaravi · x · 2026-09-22
A developer built a "Right-Click Anything" demo chaining three models now on Meta's Model API:
- Muse Spark captions the scene and extracts noun-phrases
- Those phrases prompt SAM 3.1 to predict instance masks for every object, so you can right-click any object like text on a web page
- "Look Up" hits Muse Spark web search for real product listings
- "Copy" returns an alpha-matted transparent PNG
- "Remove" paints the SAM 3.1 mask into the image and has Muse Image inpaint the object away
The author notes SAM 3.1 also supports video, leaving plenty of room for more combinations.
More from Multimodal
- Runway demos model that redesigns objects to fit new scene dimensions — _AustinCalvert_ · 2026-09-22
- Pexo Launches AI Video Agent That Lets You 'Vibe-Create' Videos via Chat — aftahi_ai · 2026-09-22
- A 27B model one-shot recreates 618 visual styles as SVG artworks, text-only — Seromelhor · 2026-09-22
- Runway Shows Video Model Understanding Scenes in 3D, Not Flat Layouts — tlakomy · 2026-09-22
- Neuralink's brain-to-voice ALS demo hits 5.4M views in 2 days, powered by Grok Voice — alvelda · 2026-09-22
- VRAM vs Blackwell Speed: 3090 Ti or 5080 for Local Video Generation? — Apollorax · 2026-09-22