Ahead of I/O, we’re releasing an updated Gemini 2.5 Pro! It’s now #1 on WebDevArena leaderboard, breaking the 1400 ELO barrier! Our most advanced coding model yet, with stronger performance on code transformation & editing. Excited to build drastic agents on top of this!
MULTIMODAL AI
-
Gen-4 References: 3D Object Positioning with AI
By
–
With Gen-4 References you can use 3D models to quickly position objects exactly where you want them. With just three images: a location, a 3D asset and a style reference, you can generate highly refined outputs without any rigging, texturing or rendering. pic.twitter.com/NQtPXGhKJd
— Runway (@runwayml) 6 mai 2025With Gen-4 References you can use 3D models to quickly position objects exactly where you want them. With just three images: a location, a 3D asset and a style reference, you can generate highly refined outputs without any rigging, texturing or rendering.
-
GPT-4o Demonstrates Advanced Image Understanding Capabilities
By
–
4o can do this: > a photo of a man wearing two collared t-shirts
-
Grok 3 demonstrates advanced PDF document generation capabilities
By
–
Grok 3's PDF rendering is insane. From a simple prompt, you can create PDF docs like research papers, resumes, invoices, business proposals, and even e-books. Here are 10 wild examples: (Prompts + demos included ↓)
-
AI Tools Reach Pro Level: Suno, Veo2, GPT-4o Highlights
By
–
Les outils d’IA ont atteint un niveau impressionnant : Suno 4.5 produit de la musique de qualité pro, les modèles de raisonnement assistent efficacement au travail, Veo2 génère des clips époustouflants, et GPT-4o crée des images natives d’un réalisme bluffant. La technologie
-
Offering help with Sanskrit OCR and VLM project
By
–
Congratulations! If you are going an OSS route, would love to help on the OCR /VLM that understands Sanskrit project!
-
Gemini now allows up to 10 file attachments on web and Android
By
–
Now you can attach up to 10 files to Gemini on web and Android. Almost BREAKING
-
Overview of Grok 3 AI Model Capabilities and Reasoning Features
By
–
Surprising Facts About Grok 3 AI Model
Boosts business efficiency dramatically
• Advanced reasoning
• Multimodal capabilities
• Real-time insights Click below to read more: https://
godofprompt.ai/blog/surprisin
g-facts-about-grok-3-ai-model
… -
California Startup Unveils π0.5 Vision-Language-Action Robotics Model
By
–
Une startup californienne annonce une percée dans la robotique à usage général avec π0.5 AI — un modèle vision-langage-action.
— VISION IA (@vision_ia) 5 mai 2025
Pas très sexy mais il semble faire le taf. pic.twitter.com/fe4kkvQEuoUne startup californienne annonce une percée dans la robotique à usage général avec π0.5 AI — un modèle vision-langage-action. Pas très sexy mais il semble faire le taf.
-
Using Runway and Luma for AI Video Generation
By
–
3. 🎥 Produisez des vidéos virales
— Jouhatsu | AI Influence Operator (@Jouhatsu_ai) 5 mai 2025
• Runway Gen-3 : Scènes cinématographiques
⁰• Luma : Mouvements fluides et dynamiques
⁰• Création et animation guidées par prompt pic.twitter.com/g9QgK6Fxmg3. Produisez des vidéos virales • Runway Gen-3 : Scènes cinématographiques
• Luma : Mouvements fluides et dynamiques
• Création et animation guidées par prompt