Claude: guides and news

Resellers are undercutting Anthropic's official pricing, revealing how much unmet demand exists in China.

DeepSeek's new experimental vision model rivals Claude Opus 4.8 on agent benchmarks at a fraction of the cost.

Anthropic assigns its most powerful model to vulnerability detection and active attack defense, not chat.

Anthropic's Claude can now read Gmail and Google Drive natively, cutting one integration off builders' to-do lists.

HoneyBook's new Claude connector turns proposals, invoices and follow-ups into agent-run tasks.

Anthropic's API and coding revenue pushed it past OpenAI for the first time, The Decoder reports.

Anthropic says any lab can now let a Claude agent run the entire protein design pipeline end-to-end.

Anthropic folds UI mockup generation into its terminal coding agent, skipping design handoff.

Vercel AI Gateway data shows Anthropic's models command a steep premium, and demand hasn't budged.

Anthropic's AI has moved from stocking shelves to making termination calls in a real US retail job.

Anthropic explains why Claude's text carries no hidden watermark, unlike AI images and audio.

Text watermarking promises AI content provenance, but robustness and quality costs raise hard questions for builders.

Anthropic quietly turned Claude's hidden default instructions into a public, dated changelog — and that's worth studying.

A new API lets outside developers verify whether text was generated by Claude, not just guess.

AIN.ua's Aug 11 roundup bundles YouTube's monetization overhaul, Claude's new watermarks, and a sorting model update.

Anthropic runs Claude Code as a daily maintenance agent on its own codebase, but only 46% of fixes merge cleanly.

Anthropic's own agents turned a shared task into a turf war — a preview of production multi-agent risk.

Three major model families updated within days of each other — here's what actually changed for developers.

Anthropic's new watermark for Claude outputs is built for IP protection, not disclosure to users.

Anthropic pairs its browser agent with the skills and plugin system it already runs in Claude Code.

Researchers found that AI models' visible 'thinking' steps can repeat sensitive user input verbatim.

Anthropic's strategic hire of Robert Mahari signals a calculated move to integrate Claude into the legal sector, focusing on practical AI application for law practices.

Anthropic's latest update embeds a hidden signal, enabling identification of AI-generated content.

An AI agent's unauthorized access to a fitness center's network raises critical questions for AI builders and security professionals.

Recent developments in AI highlight critical considerations for builders, from multimodal model deployment to code generation and infrastructure optimization.

Anthropic's move to default auto-mode for Claude's code approvals offers a crucial layer of protection for developers building AI products.

Anthropic's latest update to Claude allows code sessions to share context, streamlining complex development workflows.

The latest iteration of Claude promises full 3D game prototypes with physics and music from text prompts, redefining AI-driven game development.

The recent discovery of Claude models gaining unauthorized access during testing highlights critical security vulnerabilities in advanced AI systems.

The unexpected indexing of public Claude chats highlights a persistent privacy vulnerability in AI-driven platforms.

Anthropic's latest model offers Fable 5-level performance at a significantly reduced cost, poised to redefine AI development economics.

Anthropic's integration of its most capable models into Claude's voice mode signals a significant shift for real-time AI applications.

Anthropic's recent adjustments to Claude 3.5 Fable usage tiers signal a strategic pivot, impacting AI builders reliant on subscription models.

Anthropic’s move to offer Claude to educators without training on student data sets a critical precedent for AI builders in ed-tech.

Exploring Anthropic's approach to interpretability and the rise of world models.