JARVIS can now see. Point your camera and say "describe what you see" — it captures the live frame and sends it to Gemini's vision model, which describes the scene back in real language, spoken aloud. No key? It falls back to on-device object typing. Also supports OpenAI vision models.
Shelf Shame: upload a photo of your fridge and FridgeVision roasts your food choices, scores the fridge 0–100 with a savage verdict, and serves up one salvage recipe built from what's actually visible — then renders a shareable verdict card with the roast, score and recipe. Funny, visual, and built for screenshotting.
DaCrackFox82044
Hikima is a production-ready bilingual (English & Hausa) AI assistant. It answers clearly and professionally, asks a follow-up question when your request is unclear, says "I don't know" instead of inventing facts, detects your language automatically, and keeps conversation context. Works standalone as an offline demo, and flips to a live LLM+RAG backend (included) with one config value.
Aliyuibrahim yankwani577
Play Salty Paws — Castaway Cove — a browser mini-game.
Hassan132.8k11.4k
Play Shibaria — What's the dog doing? — a browser mini-game.
A self-contained Three.js toy-gun battle royale in a giant house. Serve this directory as the HTTP root; no build or remote runtime dependencies are needed.
Nicolas105.7k9.5k