#LLM

Anthropic Released Opus 4.8: Model Showed Equal Alignment with Claude Mythos

May 29, 2026·2 min

Mistral Released Voxtral TTS — Lightweight Text-to-Speech Model for Voice Agents

Pelicans on Bicycles: Simon Willison's Amusing Test for LLMs

Simon Willison, Django creator, invented a fun test for LLMs: asking an AI to draw an SVG pelican on a bicycle. The results reveal unexpecte

01Anthropic Released Claude Opus 4.8 with Dynamic Workflows and Cheaper Fast ModeMay 29, 2026·3 min 02Liquid AI Releases LFM2.5-8B: A Compact MoE Model with 128K ContextMay 29, 2026·2 min 03Warp Launched a Platform for Development Coordination Through AI Agents Based on GPT-5.5May 29, 2026·2 min 04Developer Created an AI-Integrator Profession: How to Build a Workflow on n8nMay 29, 2026·2 min 05Mistral AI Launches Connectors: Built-in Integrations for AI AgentsMay 29, 2026·3 min

Top themes

LLM·ZDNet AI

Why I Abandoned Copilot and Chose Claude for Word, Excel, and PowerPoint

Claude AI now works directly in Microsoft 365 and helps create, edit, and analyze documents better than the built-in Copilot.

May 29, 2026·2 min

01MUFG Becomes an AI-native Organization, Choosing ChatGPT Enterprise as Its Transformation PlatformMay 29, 2026·2 min 02How to Configure a Local Model in Ollama: Parameters for Full ControlMay 29, 2026·2 min 03Context-pruning for long-lived LLM agents: a memory management techniqueMay 29, 2026·3 min

LLM·Hugging Face Blog

IBM and Artificial Analysis create benchmark: AI agents fail at IT tasks

Large language models scored less than 50% on the new ITBench-AA benchmark for assessing AI agents' ability to solve corporate IT tasks. Thi

LLM·TNW

DeepSeek Makes 75% Discount on V4 Pro Permanent — Prices Fall 4x

DeepSeek announced that the 75% discount on its flagship V4 Pro model is becoming permanent, significantly complicating competitors' fight i

LLM·MarkTechPost

Microsoft Research releases Webwright — browser agent that solves web tasks at 60%

Microsoft Research introduced Webwright — a browser agent that executes complex web tasks better than large language models: 60% success on Odysseys benchmark versus 33.5% for baseline GPT-5.4.

LLM·MIT Technology Review

World Models: How AI Learns to Understand Reality Instead of Text

LLM·TNW

Anthropic blacklisted by Pentagon, but NSA cleared to use Claude

LLM·TNW

Claude Code Leak: Anthropic Accidentally Publishes Entire Source Code to Public npm

LLM·TechCrunch

Spotify Added AI-Generated Briefs and Q&A for Podcasts

LLM·DeepMind Blog

Google DeepMind Turned the Mouse Cursor Into a Contextual AI Assistant

StepFun Releases StepAudio 2.5 Realtime Voice Model with Roleplay Support

Langfuse for LLM Engineers: Complete Tracing and Experimentation Pipeline

How Replit Agent Solves User Loss Problem with a Single Prompt

LLM·CNews AI

AI Tools Flood Linux Maintainers with Duplicate Vulnerability Reports

Linux kernel maintainers are overwhelmed by the stream of automatic vulnerability reports generated by AI tools designed to find bugs.

LLM·Habr AI

Why RAG Chatbots Work Great in Demos but Produce Nonsense in Production

RAG bots shine on internal documentation demos but confidently generate nonsense in real work. A story about the gap between five prepared q