分类筛选
找到 1605 个 Demo
Google quietly built 9 free AI tools most people s...
Google quietly built 9 free AI tools most people still don’t know about. And the wild part? Many people are paying monthly for tools that do similar things. Here are the ones worth exploring: → Google AI Studio Build and test AI apps → NotebookLM Turn documents into a research assistant → Learn Your Way Learn topics in a more personalized way → Nano Banana Create images from text → Google Opal Turn ideas into simple apps without code → Google Pomelli Create branded social content faster → Gemini Gems Build your own custom AI assistant → Google Whisk Remix images with AI → Google Lumiere Generate video from text AI tools are getting more powerful. But the real advantage is knowing what already exists-before paying for another subscription. Bookmark this list before it disappears from your feed.
Really cool, a Gemini-powered robot! Built with G...
Really cool, a Gemini-powered robot! Built with Gemini 2.0 + OpenCV + lerobot🤗
google also seems cautious about releasing gemini...
google also seems cautious about releasing gemini 3.5 pro gemini 2.5 pro, 3.0 pro, and 3.1 pro all topped benchmarks, so google usually only ships models when they're highly competitive but right now, this is a bad time to be making headlines about your models' capabilities
This everyone should KNOW IT. GOOGLE BUILT 9 FREE...
This everyone should KNOW IT. GOOGLE BUILT 9 FREE AI TOOLS AND FORGOT TO TELL ANYONE. Most people are paying for the exact same things. ↳ Google AI Studio: build your own AI ↳ NotebookLM: any doc becomes a research assistant ↳ Learn Your Way: AI that adapts to how you learn ↳ Nano Banana: text to images, free ↳ Google Opal: idea to working app, no code ↳ Google Pomelli: brand to social content, automatic ↳ Gemini Gems: your own personal AI assistant ↳ Google Whisk: remix any image with AI ↳ Google Lumiere: text to video, free Midjourney: $10/month. Runway: $15/month. OpenAI: $20/month. Google built all of this. Charged you nothing. Told nobody. Bookmark this before your feed buries it.
HERMES AGENT MIXTURE OF AGENTS COMBINES MULTIPLE M...
HERMES AGENT MIXTURE OF AGENTS COMBINES MULTIPLE MODELS INTO ONE ANSWER. 8% HIGHER THAN OPUS 4.8. 11% HIGHER THAN GPT-5.5. NO GATED ACCESS REQUIRED. Mixture of Agents (MoA) runs multiple models on the same query in parallel. an aggregator model synthesizes all responses into one answer that outperforms any single model alone. Nous Research benchmarks: 8% higher than Opus 4.8. 11% higher than GPT-5.5. (upcoming benchmark, numbers from official announcement.) HOW IT WORKS: you select a MoA preset as your model. Hermes fans out your query to 2-3 reference models. each responds independently. the aggregator reads all responses and synthesizes one final answer. you see one response. behind it: multiple perspectives. DEFAULT PRESET: moa: default_preset: default presets: default: reference_models: - provider: openai-codex model: gpt-5.5 - provider: openrouter model: deepseek/deepseek-v4-pro aggregator: provider: openrouter model: anthropic/claude-opus-4.8 reference_temperature: 0.6 aggregator_temperature: 0.4 max_tokens: 4096 enabled: true two reference models generate diverse responses. Opus aggregates them into one answer. the output is better than any of the three alone. SETUP: Desktop app / Dashboard → Models → MoA presets CLI: hermes moa configure create named presets for different tasks: hermes moa configure # default preset hermes moa configure review # create "review" preset hermes moa configure research # create "research" preset hermes moa list # see all presets hermes moa delete review # remove a preset BUILD YOUR OWN PRESETS: CHEAP RESEARCH (2 models, budget): presets: research_lite: reference_models: - provider: openrouter model: deepseek/deepseek-v4 - provider: openrouter model: google/gemini-2.5-flash aggregator: provider: openrouter model: anthropic/claude-sonnet-4.6 diverse perspectives at budget prices. Sonnet aggregates. good enough for daily research. MAXIMUM QUALITY (3 models, premium): presets: full_power: reference_models: - provider: openai-codex model: gpt-5.5 - provider: openrouter model: deepseek/deepseek-v4-pro - provider: openrouter model: google/gemini-2.5-pro aggregator: provider: openrouter model: anthropic/claude-opus-4.8 three frontier models + Opus aggregation. this is the preset that beats benchmarks. MID-SESSION TOGGLE: /moa # toggle MoA on/off /moa research # switch to named preset /moa off # disable, use plain model working on routine code? MoA off. hit a hard architectural problem? /moa on. one command. no config edit. no restart. SAFETY RAILS: → aggregator cannot be another MoA preset (recursive MoA trees blocked) → enabled: false disables fan-out (aggregator acts as a plain model) → each reference model runs in parallel (wall clock ≈ slowest model, not sum) WHERE MoA MAKES SENSE: → complex architecture decisions → research synthesis across diverse sources → code review where one model misses edge cases → critical content that needs multi-perspective verification → any task where "second opinion" matters WHERE MoA IS OVERKILL: → simple file edits → routine web searches → cron jobs and monitoring → tasks where speed matters more than depth MoA multiplies your token cost by the number of reference models. use it for the 10% of tasks where quality matters most. full Hermes architecture deep-dive in the article 👇
Gemini 2.5 Pro is now available for scaled paid us...
Gemini 2.5 Pro is now available for scaled paid usage and is going to preview today! If you want to keep using Gemini 2.5 Pro on the free tier, keep using the experimental version (no change needed), both are the same model under the hood. See details in 🧵 https://t.co/wilPQDCc4x
🚀 Stop paying for AI tools when some of the most...
🚀 Stop paying for AI tools when some of the most powerful ones are already FREE. Most people only know a few AI tools… but Google quietly built an entire AI ecosystem that can automate your work, content, coding, and productivity. Here are 11 powerful free AI tools from Google that can completely upgrade your workflow:📈 • Google Gemini – Your AI assistant for research, writing, and problem solving. • Google Spreadsheet – Generate formulas, insights, and automate data analysis instantly. • Google AI Studio – Test and experiment with advanced AI models and prompts. • Notebook LM – Turn documents into summaries, mind maps, and AI-powered learning insights. • Google Gemini – Host real-time AI conversations with screen sharing for meetings or presentations. • Google App Builder – Create apps quickly without heavy coding. • Google Firebase Studio – Build and launch AI-powered apps faster. • Veo – Generate videos from simple prompts. • Google Gemini Nano Banana – AI image editing and creative visual generation. • Imagen – Create high-quality AI images from text prompts. • Google Gemini Ask on YouTube – Chat with YouTube videos to get quick insights and summaries. 💡 The biggest opportunity right now isn’t just learning AI — it’s learning how to use AI tools before everyone else does.
毎週水曜日の #Gemini3 練習会 メニュー： ・200m × 5（計測ミスで記録上は 4本）...
毎週水曜日の #Gemini3 練習会 メニュー： ・200m × 5（計測ミスで記録上は 4本） ・6km テンポラン フルマラソンに向けた基礎練習を開始 結果はいずれも 32秒台… もう少しスピードを出せるようにしたい 次のフルマラソンまで残り 4か月を切っている…！ #GeminiRunners #まるお製作所RC #ジェミラン
Google ve OpenAI Aynı Günde Model Güncellemelerini...
Google ve OpenAI Aynı Günde Model Güncellemelerini Duyurdu Google Gemini 3.5 Flash modelinde bilgisayar kullanımını artık yerleşik bir araç olarak sunduğunu açıkladı. Daha önce ayrı bir Gemini 2.5 modeli ile sunulan bu özellik ana Flash modeline entegre edildi. Model zaten işlev çağrısı, arama ve haritalama araçlarında güçlüydü şimdi tarayıcı, mobil ve masaüstü ortamlarında görebilen, akıl yürütebilen ve eylem alabilen ajanlar oluşturmak mümkün hale geldi. Bu sayede uzun süreli kurumsal otomasyon görevlerinde örneğin sürekli yazılım testi veya profesyonel uygulamalarda bilgi işlerinde daha iyi sonuç alınıyor. Google, prompt injection risklerini azaltmak için özel adversaral eğitim uyguladı ve isteğe bağlı iki güvenlik sistemi ekledi hassas eylemler için kullanıcı onayı ve dolaylı prompt injection tespitinde görevi otomatik durdurma. Browserbase Browser Use ve UiPath gibi şirketler bu modelle frontier seviye modellere yakın doğruluk elde ederken daha iyi maliyet ve hız performansı gördüklerini belirtiyor. Gönderi: https://t.co/tVszadqjFJ OpenAI ise en çok kullanılan modeli GPT-5.5 Instant’ı güncelledi. Yeni sürüm sorunun arkasındaki amacı daha iyi anlıyor ve buna göre yanıtını uyarlıyor. Karmaşık kısıtlamaları daha güvenilir şekilde ele alıyor, alışveriş ve yerel önerilerde daha tutarlı ve faydalı sonuçlar veriyor. Güncelleme bugün ücretli kullanıcılara, yarın ise ücretsiz kullanıcılara açılacak. Bu iki gelişme birlikte hem ajan tabanlı bilgisayar otomasyonunda hem de günlük sohbet ve öneri kalitesinde somut ilerlemeler sağlıyor ve geliştiricilerin daha pratik uygulamalar üretmesine katkı sunuyor. Gönderi: https://t.co/PFvHkEZ9gf Google gelişmesini ücretsiz deneyebiliyoruz. Denemek lazım ne kadar iyi görmek lazım. İkisini birlikte atayım dedim güzel gelişmeler. Reklam değildir.
Google's former CEO just revealed what industry in...
Google's former CEO just revealed what industry insiders have been saying all along Wealth in AI isn't built by scrolling through hype It's built by mastering the actual tools Agents, Interactions API, prompts, memory, skills, MCP, and automation workflows Your competition is already learning this Everything below is free: Gemini basics https://t.co/Vg3JVFDhPP Gemini quickstart https://t.co/Z82wKSlTK2 Prompt design strategies https://t.co/5UZ3sAdmXl Gemini API reference https://t.co/0twEsqIAAd Models overview https://t.co/FlkQOk8s73 Google AI Studio https://t.co/7MquFNlJtm Interactions API overview https://t.co/I3qGjONi9b Interactions API quickstart https://t.co/KSQRsmYSvJ Agents overview https://t.co/YVRsLRKaBi Antigravity Agent https://t.co/wOoFNSimWh Function calling guide https://t.co/xVnAL04Oa2 Using Tools with Gemini API https://t.co/PZ7JfaRA8w Structured outputs https://t.co/L9UXnuvD1g Caching https://t.co/W7VhqfT9CV Model Context Protocol https://t.co/E2UvIQF6vg Multimodal Live API https://t.co/SyBHNSEF0I Google Gemini cookbook https://t.co/d2tNGBmewu System instructions https://t.co/sEqDoWzFGU All of this costs $0 Most people will keep asking AI one question at a time Save this and start learning the stack
Google DeepMind just dropped Gemma 3 and it’s a bi...
Google DeepMind just dropped Gemma 3 and it’s a big deal. This is the most powerful AI you can run on a single GPU or TPU. That means fast, efficient AI that works on everything from laptops to high-end workstations, no need for expensive cloud setups. Here’s why everyone is talking about Gemma 3 ⬇️ Why Gemma 3 is a Game Changer Gemma 3 is based on the same tech behind Gemini 2.0, but designed to be lighter, faster, and more accessible for anyone building AI applications. ✅ Super powerful on a single GPU – Beats other models like Llama3-405B and DeepSeek-V3 in real-world tests. 🌍 Understands 140 languages – Build AI that speaks to the whole world. 📝 Handles massive text inputs – 128K context window lets it process long conversations and documents easily. 📷 Works with images & videos – Analyze pictures, short videos, and text all in one model. ⚡ Fast & efficient – Uses less power and storage while still delivering top-tier performance. It’s about making AI more accessible to everyone. How Does It Compare? Most AI models this powerful require huge cloud servers with dozens of GPUs. But Gemma 3 27B runs on just ONE GPU. For comparison: 💥 Mistral Large & Llama 3 405B need up to 32 GPUs to reach similar performance. This means lower costs, more flexibility, and better access for developers and businesses. Safety First: Built to Be Responsible AI is evolving fast, and safety is crucial. 🔒 Strict safety testing – To prevent misuse. ✅ Fine-tuned for safe responses – To align with ethical AI guidelines. 🛡️ ShieldGemma 2 – A built-in AI tool that filters harmful or unsafe image content in real-time. Here’s How to get started: 💻 Try it instantly – Run it in Google AI Studio (no setup required). 📥 Download it – Available on Hugging Face, Kaggle, and Ollama. 🛠️ Customize & fine-tune – Works with PyTorch, JAX, UnSloth, and more. ☁️ Deploy anywhere – Use Vertex AI, Cloud Run, or NVIDIA GPUs. 🎓 For researchers – Google is offering $10,000 in Cloud credits for academic projects. With 100M+ downloads and 60K+ community-created models, Gemma 3 is part of a massive open-source AI movement.
2週間ぶりの #ジェミラン 200×5 その後のペーランはキロ8で心拍160でした😭 早く心肺を戻...
2週間ぶりの #ジェミラン 200×5 その後のペーランはキロ8で心拍160でした😭 早く心肺を戻したいところ #まるお製作所RC #Gemini3 https://t.co/CcZy0fK36G