OpenAI just launched the GPT-5.2 model family—spanning Instant, Thinking, and Pro tiers—built specifically for professional workflows. The Thinking variant cuts hallucinations by 30% relative to GPT-5.1 and matches or beats human professionals on 70.9% of knowledge tasks across 44 occupations. It is rolling out to paid ChatGPT subscribers and API developers today.
Orchids just launched what it calls the world's first vibe coding IDE, claiming the #1 spot on the end-to-end software development benchmark App Bench. The tool packs an AI agent, code editor, built-in browser, Supabase, and Stripe directly into a local environment. Why it matters: It tries to push vibe coding beyond toy frontends by putting full-stack infrastructure into a single workflow.
Mistral AI just released Devstral 2, a lightweight open-weights coding model family led by a 123B flagship and a 24B small version. The larger model scores 72.2% on SWE-bench Verified at up to 7x the cost efficiency of Claude Sonnet, while the 24B version runs locally on consumer hardware. The release also includes Mistral Vibe, an open-source terminal CLI for autonomous software development.
Accenture and Anthropic just announced a major partnership expansion, forming a dedicated business group and training roughly 30,000 consultants on Claude. The consulting giant is also rolling out Claude Code to tens of thousands of its developers, aiming directly at regulated industries like finance and healthcare to move AI experiments into production.
Alibaba's open-source Qwen Code CLI just rolled out updates spanning v0.2.2 to v0.3.0, introducing Stream JSON input and output. That makes hooking the coding agent into CI/CD pipelines, SDKs, and automated workflows far cleaner, alongside full UI localization and a 20MB buffer guardrail against memory crashes.
Anthropic just integrated its Claude Code assistant directly into Slack as a research preview. Instead of copying and pasting bug reports into a separate tool, developers can now simply @Claude in a chat thread to pull context and linked repositories for quick debugging. You'll just need to connect your repos via the web version first.
Google just bumped up rate limits for Antigravity, its new agentic development platform. Pro and Ultra subscribers now get top-tier priority access with quotas refreshing every five hours, while free users switch to larger weekly caps. All tiers retain access to Gemini 3 Pro and unlimited tab completions.
Google just launched Workspace Studio, letting any employee build and share custom AI agents without writing code. Powered by Gemini 3, these agents can reason through complex tasks like prioritizing emails and connecting workflows across Gmail, Drive, and Salesforce. Why it matters: It turns everyday office workers into workflow automators in minutes.
Krea AI just launched Node App Builder, a tool that lets users convert complex workflows into functional apps without writing code. Built on a graph architecture supporting up to 1,000 nodes per app, AI automatically optimizes step connections. It basically turns internal team automations into standalone, shareable software.
Anthropic just announced the acquisition of Bun, the fast JavaScript runtime and toolkit, to bolster its coding infrastructure. The deal comes as Claude Code reached a massive $1 billion run-rate revenue milestone just six months after launch. Bun will remain open source under the MIT license while supercharging Claude Code workflows.
AWS just launched three new "frontier agents" in preview—Kiro autonomous agent, AWS Security Agent, and AWS DevOps Agent. Designed to run autonomously for hours or days, they handle everything from multi-repo coding and on-demand penetration testing to diagnosing infrastructure outages in minutes. Basically, Amazon wants AI to act like full virtual teammates rather than just passive copilot assistants.
Hugging Face just rolled out Agent Skills, standardized instruction packages that teach coding agents like Claude Code, Cursor, and Codex how to train models, manage datasets, and run evals. Each capability loads on demand through a simple SKILL.md file, basically turning generic coding assistants into full-stack ML engineers across the HF ecosystem.
Anthropic just published an internal study examining how tools like Claude Code are reshaping its own engineering teams. Employees report using Claude across roughly 60% of their work to gain an average 50% productivity boost, tackling projects that previously weren't worth the manual effort. Why it matters: Even the people building frontier AI share the same anxieties, reporting concerns over skill atrophy and reduced mentorship from peers.
Anthropic just released Claude Opus 4.5, calling it the world's best model for coding, agents, and computer use. It even outperformed human candidates on the company's internal software engineering exam. With API pricing lowered to $5/$25 per million tokens and a new effort parameter to cut token usage, frontier intelligence is getting both smarter and cheaper.
Hugging Face just launched AnyLanguageModel, an open-source Swift package that serves as a drop-in replacement for Apple's Foundation Models framework. By swapping a single import statement, developers can alternate between local engines like MLX and Core ML and cloud providers without rewriting glue code. It dramatically lowers the bar for running open-source models on Apple silicon.
OpenAI just published a playbook for business leaders on turning messy AI pilots into reliable production workflows using contextual evals. The framework details how cross-functional teams should build golden test sets, blend LLM graders with human review, and spin up a continuous data flywheel. It basically argues that as foundation models commoditize, proprietary evaluation criteria will become the real enterprise moat.
OpenAI just launched GPT-5.1-Codex-Max, an agentic coding model built to run continuously for over 24 hours. Using native context compaction, it works seamlessly across multiple windows on multi-million-token refactors while cutting thinking tokens by 30%. It is now the default engine inside Codex.
Replit just launched Design Mode, powered by the new Gemini 3 model to turn plain English prompts into polished, interactive mockups and static sites in under two minutes. You can preview clickable layouts right away and hit Convert to App whenever you're ready to add real backend code.
Microsoft is doubling down on its flagship Earth-system AI model, Aurora. In partnership with the University of Cambridge, the company pledged to keep future releases and full training pipelines completely open-source, aiming to help local weather services and energy grids build tailored forecasting tools with minimal data.
Microsoft just rolled out GPT-5.1 as an experimental model in Copilot Studio for U.S. customers in early-release Power Platform environments. The new model offers adaptable thinking time across chat and reasoning tasks, though Microsoft recommends keeping it out of production for now.
