AI Tools news and analysis on Web Pulse: 38968 curated articles tracking the latest AI Tools developments, tools, and research, updated continuously from vetted sources.
Ollama released version 0.34 on September 5, adding macOS integration that lets ChatGPT Desktop run local Ollama models such as Gemma 4 and Llama, plus a new cloud tier for large open-weight models. The local integration…
A developer's attempt to render AI agent state (working, blocked, idle) directly into the tmux window list failed on both candidate signals: Claude Code sets its pane title once and never updates it, and tmux's pane_curr…
China Mobile has open-sourced Open-RAIL, an engineering base that connects vision-language-action (VLA) and world-action-model (WAM) systems with robot bodies, supporting four heterogeneous robots and 10 VLA or WAM model…
A NixOS user reported that Anthropic's Claude AI fixed an endless Stardew Valley online loading screen by patching the headers of two GOG Galaxy SDK binaries, libGalaxy64.so and libGalaxyCSharpGlue.so, so they no longer …
The Model Context Protocol released its 2026-07-28 version, formally locking in the MCP Apps extension that lets servers return isolated, interactive, stateful UIs embedded inline with agent responses. MCP Apps were prop…
A developer-authored explainer outlines what AI agents are and why companies are adopting them, defining an agent as an AI system that perceives its environment, plans, takes actions, and retains memory to complete goals…
Prior Labs released TabPFN-3.5, a tabular foundation model that scores 0.375 on the private leaderboard of the 2015 Otto Group Product Classification Challenge, beating the winning solution's 0.382, according to Prior La…
Salesforce announced Koa, a domain-specific reasoning model built by post-training Nvidia's open Nemotron 3 Super model on synthetic data from nearly three decades of business knowledge, at its Dreamforce 2026 event in S…
A developer published a reusable prompt that instructs a coding agent to build a production-quality admin insights and operations dashboard inside an existing repository, adapting to the project's stack and business mode…
The gptimage25.ai project released an independent, open-source GPT Image 2.5 prompt collection containing 30 illustrated cases plus 20 original editing recipes, for 50 copyable prompts total, each in English and Chinese.…
A developer published a step-by-step tutorial showing how to evolve a customer-support assistant from deterministic rule-based code into a full agentic AI system using Python. The progression moves through ten stages, in…
A developer outlined a five-step method for building a campus IT helpdesk chatbot using Tencent EdgeOne Makers, an integrated AI platform that lets institutions deploy AI agents without standing up their own server or AI…
RomeAILab has released Rome, an early-stage open-source "agentic OS" written in TypeScript that aims to let AI agents accumulate state, memory, and self-improvement across runs rather than starting from zero each convers…
A developer has published a step-by-step guide for migrating stateful agent logic from n8n prototypes into LangGraph production agents, arguing that only the orchestration layer needs to move rather than the entire workf…
OpenAI released GPT-Live-1 into its API on September 10, a full-duplex voice model priced at $0.05 per minute for the voice layer, with backend reasoning billed separately at the chosen model's own rate. Real-world total…
Huawei has opened beta testing for XiaoYi Work, an AI work assistant for office tasks, coding and creative work across HarmonyOS phones, tablets and PCs. XiaoYi Work can break a goal into steps, call tools, show progress…
AI-powered bug hunting has driven a surge in vulnerability disclosures in 2026, according to Gartner research vice president Craig Lawson, who said codebases are being audited at a level never seen before. The expanded a…
Vivo previewed BlueCode, a phone-based coding agent for AI-assisted software development on mobile devices, at vivo Developer Conference 2026. The company also said it is exploring a 30-billion-parameter BlueLM MoE model…
A full stack developer compared Claude Code, Codex, and Cursor, concluding that no single AI coding tool is universally fastest and that each fits a different part of the development loop. The developer recommends Cursor…
A developer published a guide on tamiz.pro arguing that AI-generated tests should be treated as untrusted input and gated through four SDLC checks before merging. The proposed pipeline enforces test execution, coverage t…