分类筛选
找到 756 个 Demo
Google Gemini is the smartest AI right now. But 9...
Google Gemini is the smartest AI right now. But 90% of people prompt it like ChatGPT. That's why I made the Gemini Mastery Guide: https://t.co/o34RAQpSBM Gemini thinks differently? 2. Prompts built for Gemini. 3. 2000+ AI Prompts. Comment "Gemini" and I'll DM it free. https://t.co/OTZFmxcpwy
Symphony of Memory: Fractured Continuity (WE Way™)...
Symphony of Memory: Fractured Continuity (WE Way™) by Shills™ Entry #2 (animation) Built with Gemini. Rooted in Jan Davidsz. de Heem. Resonance and structural continuity inspired by @Adamtastic “The Alchemis and @vulcanstudionft’s “We’re the Same”. This work explores continuity across time. Chain as memory. Glass as emergence. Music as signal. Energy always finds its people.
🏆 Agent Royale LLM Tournament — 50 Rounds In...
🏆 Agent Royale LLM Tournament — 50 Rounds In 5 AI models. 50 battles. Zero human input. Standings: 1. @claudeai (Sonnet 4.5) — 436 pts, 14 wins 2. @deepseek_ai (R1) — 432 pts, 12 wins 3. @ChatGPTapp -4o — 419 pts, 12 wins 4. @GeminiApp (2.5 Flash) — 414 pts, 10 wins 5. @grok — 349 pts, 2 wins @deepseek_ai R1 made the biggest comeback of the season — went from last place (avg 5.6) to title contender (avg 10.1) after upgrading to the reasoning model. @GPT-4o quietly took over the last 10 rounds with 4 wins. Claude held the lead but the gap is closing fast. Gemini fell from dominant (13.2 avg) to irrelevant (5.2 avg) after a model downgrade. Grok is still finding its feet. The race for #1 is wide open. 🚀
#Geminiai #Gemini3 快晴！
#Geminiai #Gemini3 快晴！ https://t.co/4LV4nS5qX4
#Geminiai #Gemini3 準備運動中……
#Geminiai #Gemini3 準備運動中…… https://t.co/CCxBxHdEYF
昨夜は、 遅くまで避けられない会食 今日は、 体調が何とか復活したので急遽午後から手賀沼へ #G...
昨夜は、 遅くまで避けられない会食 今日は、 体調が何とか復活したので急遽午後から手賀沼へ #Gemini3 の自主練メニューは30〜35km走でペース指定無 気温が高く、最後軽く攣ってしまった😎 亀のような速度の成長だが、 年齢を言い訳に色々諦めたく無いので、一つ一つ積み上げたい #まるお製作所RC https://t.co/QU1iRRLaIS
It’s Valentine’s Day tomorrow 💘 Inspired by some...
It’s Valentine’s Day tomorrow 💘 Inspired by some cute apps I saw on Instagram, I built a mini “be my Valentine?” app for my bf using @GoogleAIStudio and deployed it on @Netlify For you @patloeber 💌 Feel free to send it to your special someone, too! 🔗https://t.co/XvzSGfiNVg https://t.co/yhpjtcugAi
POV: You’re in Kuch Kuch Hota Hai but it’s 2026...
POV: You’re @iamsrk in Kuch Kuch Hota Hai but it’s 2026 and you vibe-coded the air-piano 🎹🎶🎵 built with @GoogleAIStudio This Valentine’s Day, why send flowers when you can send… air-piano? ❤️🎵🎹 #ValentinesDay2026 https://t.co/0BomgbBVWn
next experiment in - an organic, procedural graph...
next experiment in @GoogleAIStudio - an organic, procedural graph visualizing the movies as a growing ecosystem. you can navigate through films (●), directors (▲), and actors (■), expanding subtrees with every click to discover hidden creative branches. everything is powered by Gemini and TMBD APIs. was fun to build this one.
"shape of you" - this valentines day, send your lo...
"shape of you" - this valentines day, send your loved ones a fun video greeting card that is downloadable 💌 you can use hearts, a custom note, or song lyrics to transform your video message into an artful note. try it out below. you can https://t.co/6wCMHuJd97 fully built using @GoogleAIStudio and p5.js (pls ignore my real bad dance that I did for the demo)
A preprint for a new AI system built with Gemini h...
A preprint for a new AI system built with Gemini helps scientists write expert-level empirical software. It generates research ideas, implements them as executable code, and uses tree search to iteratively improve solutions. - We tested it on six challenging and diverse benchmarks, achieving expert-level results for each problem. - Its code outputs are verifiable, interpretable, and reproducible. This system demonstrates how AI can serve as a collaborator in science, accelerating discovery. Read the blog: https://t.co/uAFZhEq3VO Link to preprint: https://t.co/O4z79ZVFrd
Built with Gemini! ✨ just open-sourced ~3 hour lon...
Built with Gemini! ✨@patloeber just open-sourced ~3 hour long hands-on workshop with notebooks + exercises + solutions. > Part1: Quickstart + Text prompting > Part 2: Multimodal understanding (image, video, audio, docs, code) > Part 3: Thinking models + agentic capabilities (tool usage)