Generative AI news and analysis on Web Pulse: 15422 curated articles tracking the latest Generative AI developments, tools, and research, updated continuously from vetted sources.
Hedra said it cut 2K image edits with Krea 3 to 3.6 seconds on AMD Instinct MI355X accelerators, down from about 22 seconds on the same hardware, after rebuilding its inference stack with custom ROCm kernels and lower-pr…
OpenAI said on October 15th that GPT-Live-1.1, its real-time voice system, is rolling out to ChatGPT, Codex and Dots with new voices in French, German, Hindi, Mexican Spanish, Korean and Japanese. The announcement on X a…
Krea launched Krea 3 in open beta, an AI image editing model that completes native 2K-resolution edits in under 8 seconds at $0.06 per image — 3.7x cheaper and 11x faster than GPT Image 2 ($0.22, 90 seconds median) — ava…
Bluesky announced at TechCrunch Disrupt 2026 that it is expanding its AI-powered Attie platform so everyday users can build websites, quizzes, games, and polls that run on the AT Protocol (atproto). Bluesky CEO Toni Schn…
Black Forest Labs released FLUX 3 Image Klein on October 15th, a 5.6-billion-parameter image-generation and editing model that the Freiburg-based lab says runs on consumer GPUs, including 8GB-VRAM hardware in quantized v…
Meta released Muse Spark Open, an open-weight model with 816 billion parameters that handles text, images, video, and audio with up to one million tokens of context, positioning it against Chinese open-weight models Qwen…
Microsoft launched Copilot for Legal on October 22, 2026, adding more than 20 legal-specific skills to Microsoft Copilot in Word and its agentic workspace Cowork, plus plugins from more than two dozen legal technology pa…
A non-CS student built a specialized menu translator for a local restaurant that handles dish-specific context better than generic tools like Google Translate. The web app uses a JSON structure to encode how each dish is…
Meta Platforms released Muse Glimmer, an open-weight AI model with approximately 30 billion parameters that runs locally on a Mac or a PC with a single consumer graphics card, on August 10, 2026, while keeping its more a…
A developer built AI Zombie Couple Video, a Next.js web app that turns two portrait photos into a personalized 15-second zombie love story using Seedance 2.0 or 2.5 at 480p, 720p, or 1080p, sold through one-time credit p…
Epic Games released a public preview of a rebuilt Epic Games Launcher on October 14, 2026, claiming a 5.47-second cold start versus 10.07 seconds for Steam, a 46% improvement Epic attributes to a rewrite built with AI to…
Nvidia Corp. launched the Nemotron 3.5 Super VL, a 120-billion-parameter open multimodal large language model, alongside the edge-focused Nemotron 3.5 Bolt models for agentic systems. Nvidia said Super VL matches its lar…
A new Claude Code skill automates the path from iOS app code to App Store submission, handling review-rule checks, product page copy, screenshot generation, build uploads, App Store Connect entry, submission and release …
A developer who used GrokBot to orchestrate 16 always-on agents merged 78 pull requests in a week, more than double their previous best of 35, but the setup collapsed after the bots burned $100 over a weekend and hit the…
NVIDIA released Nemotron 3.5 Super VL on October 15th, an open multimodal model for AI agents with 120 billion total parameters and 12 billion active per token that processes text, images, documents and video. NVIDIA sai…
Visko and Google Cloud TPU engineers ported Visko's Orbis Live World Model pipeline to a single Cloud TPU slice in 30 days, rendering end-to-end at 1.07x playback speed, according to Google Cloud Head of Developer Advoca…
Black Forest Labs launched FLUX 3 Image Klein on October 15, 2026, releasing the model weights for a 5.6 billion parameter text-to-image and image editing model that is free for commercial use by businesses and individua…
An independent head-to-head test of two 3D model generation tools found that Tripo AI charged 55 credits for a skull ornament generation while SupaVoxel charged 3, yet the difference in finished resin cost was only 38 ce…
A developer built MarketMitra-AI, a Flask-based social media post generator that runs the Llama 3.2 3B model locally through Ollama instead of a hosted inference API, producing captions, hashtags, calls to action and Wha…
A developer built SideHustle Listing AI, a web application that turns short product descriptions into structured e-commerce listing content such as titles, descriptions, selling points and keywords. The tool pairs a Reac…