# AI Catchup

> Catch up on AI: tools, practices, and what matters this week

A weekly newsletter and content hub covering AI tools, best practices, and the developments that actually matter.

AI Catchup publishes sourced AI news, practical guides, comparisons, tutorials, editor's picks, and a weekly newsletter. Article pages include canonical URLs, direct-answer summaries, publish and update dates, author information, primary-source links, and structured data.

## Explore

- [Tools & Software](https://aicatchup.com/tools): In-depth guides to individual AI tools and features: what each one does, how to set it up, and when to use it
- [Practices & Workflows](https://aicatchup.com/practices): How people use AI day-to-day. Prompt engineering, skill files, automation patterns
- [News & Analysis](https://aicatchup.com/news): Curated takes on AI industry news with perspective
- [Skills & Learning](https://aicatchup.com/skills): Tutorials and guides for leveling up with AI tools
- [Comparisons & Roundups](https://aicatchup.com/comparisons): Head-to-head comparisons of AI coding agents, including Claude Code, Codex, and Cursor, on architecture, parallel work, scheduled agents, and computer use
- [Editor's Picks](https://aicatchup.com/picks): Current ranked recommendations
- [Newsletter Archive](https://aicatchup.com/newsletter): Weekly AI Catchup editions
- [Search](https://aicatchup.com/search): Search the complete content archive
- [Topic Index](https://aicatchup.com/topics): Browse articles by product, model, workflow, and technology
- [Plans, Prices and Limits Tracker](https://aicatchup.com/tracker): Current plans, prices, usage limits and models per vendor, each row quoting the vendor's page with its last-checked date

## Plans, Prices and Limits Trackers

- [Cursor plans, prices and limits](https://aicatchup.com/tracker/cursor): Cursor is sold as part of The X Subscription: Free, Plus ($8 a month), Premium ($30), Super ($100) and Ultra ($200), plus per-seat Teams from $30. Each paid plan includes one weekly usage pool for Cursor's own models, shared with Grok. Third-party models bill at API list price from prepaid credit on Premium and above, plus $0.25 per million tokens on Teams.
- [Claude Code plans, prices and limits](https://aicatchup.com/tracker/claude-code): Claude Code is included in every paid Claude plan: Pro costs $20 a month ($17 a month billed annually), Max 5x $100, Max 20x $200, Team seats $20 to $125, and Enterprise $20 a seat (20-seat minimum) plus usage at API rates. Pro, Max and Team usage draws on five-hour and weekly limits shared with Claude chat.
- [OpenAI Codex plans, prices and limits](https://aicatchup.com/tracker/codex): Codex comes with every ChatGPT plan. Free ($0) and Go ($8 a month) get GPT-6 Luna in the desktop app, Plus costs $20 a month, Pro $100, $200 or $500, and Business $20 per user billed annually. Plus and Standard Business usage is estimated per five-hour period, Pro has no five-hour limit, and credits extend usage.

## Featured Evergreen Guides

- [Ramp Glass Playbook: Internal AI Workspaces for Company-Wide AI Adoption](https://aicatchup.com/practices/internal-ai-workspaces-playbook): Glass is Ramp's internal AI productivity suite, built after the company had already hit 99 percent AI adoption and found most employees still stuck on setup. One Okta SSO sign-in wires up every tool, a marketplace called Dojo holds 350+ shared skills, memory is mined from Slack, Notion, Linear and Calendar, and jobs run on cron. Here is the playbook other companies can borrow.
- [Prompt Engineering in 2026: The Playbook That Works Across Claude and GPT](https://aicatchup.com/practices/prompt-engineering-guide): Prompt engineering in 2026 has two layers: cross-model fundamentals (clear direct instructions, motivated rules, 3-5 structured examples, XML boundaries, long documents before the question) and model-specific overrides that matter more than ever, because the newest models need instructions deleted as often as added. Start here, then apply the guide for your model.
- [How to Build Your Own MCP Server: A Beginner's Guide](https://aicatchup.com/skills/building-mcp-servers): To build an MCP server in TypeScript, install the v2 SDK package @modelcontextprotocol/server with Zod v4, register each tool with server.registerTool and a Zod input schema, serve it over stdio, then add it to Claude Code with claude mcp add or a .mcp.json entry. A one-tool server is about 50 lines.
- [Codex CLI vs Claude Code vs Cursor: 2026 Architecture Deep-Dive (Sandboxing, Context, Plugins, Scheduling)](https://aicatchup.com/comparisons/codex-cli-vs-claude-code-vs-cursor-architecture): Codex CLI, Claude Code and Cursor now all confine local shell commands in an OS-enforced sandbox and all run hook scripts, so the real split is defaults and composition. Codex sandboxes by default, Claude Code's sandbox is opt-in, and Cursor sandboxes under Auto-review. Choose by default posture, Windows support and how unattended work runs.
- [Perplexity Personal Computer vs OpenAI Codex Computer Use vs Claude Computer Use: Which AI Should Run Your Mac?](https://aicatchup.com/comparisons/perplexity-vs-codex-vs-claude-computer-use): Pick Perplexity Personal Computer for an always-on desktop agent on Mac or Windows; Max includes 10,000 Computer credits a month. Pick Codex Computer Use if you live in ChatGPT and want one agent driving code, browser and desktop apps. Pick Claude Computer Use to wire it into your own harness on any OS, or in Claude Code.
- [Gemini in 2026: Google's Full AI Stack Explained (Gemini 4 Argon, Antigravity, Chrome, Workspace, AI Mode)](https://aicatchup.com/tools/gemini-google-ai-stack-2026): Google's AI stack now runs on Gemini 4 Argon, its frontier model since October 13, 2026, with Gemini 3.8 Flash as the low-cost default. Argon costs $2 input and $10 output per million API tokens through November 30, then $4 and $20. Google AI Pro and Ultra subscribers get it in the Gemini app, AI Mode and Antigravity.

## Latest Articles

- [Gemini in 2026: Google's Full AI Stack Explained (Gemini 4 Argon, Antigravity, Chrome, Workspace, AI Mode)](https://aicatchup.com/tools/gemini-google-ai-stack-2026): Google's AI stack now runs on Gemini 4 Argon, its frontier model since October 13, 2026, with Gemini 3.8 Flash as the low-cost default. Argon costs $2 input and $10 output per million API tokens through November 30, then $4 and $20. Google AI Pro and Ultra subscribers get it in the Gemini app, AI Mode and Antigravity.
- [Claude Code Auto Mode Is Default on Every Plan and Expanding to Most Apps](https://aicatchup.com/news/claude-code-auto-mode-default-august-2026): Anthropic says Auto is rolling out next week as the default for new sessions in most Claude apps on Pro, Max, and Team. Enterprise administrators will receive an email before their organization's default changes. This article also summarizes Claude Code's defaults by version, provider, and session type.
- [Claude Duet: Anthropic's Proactive Agent App, Who Gets It, and How It Compares to ChatGPT dots](https://aicatchup.com/news/claude-duet-proactive-agent): Claude Duet is a research preview of a proactive agent app from Anthropic, running on Claude Opus 5.5 in its own cloud computer. It sends a morning brief, checks connected apps every few hours, and asks before it sends, pays, signs, or deletes. Access is by waitlist for US Pro and Max subscribers, inside existing usage limits.
- [OpenAI Adds Prompt Presets and Inline Controls to Codex CLI](https://aicatchup.com/news/openai-codex-cli-presets-inline-settings-enterprise-mcp): OpenAI's Codex CLI 0.164.0 release adds reusable prompts in ten numbered `/presets` slots, inline settings for model, reasoning effort, speed, and permissions, plus sign-in options for enterprise MCP connections.
- [OpenAI Launches dots: Always-On Agents in ChatGPT With Their Own Cloud Computer](https://aicatchup.com/news/openai-chatgpt-dots-always-on-agents): OpenAI's dots are always-on agents in ChatGPT, powered by GPT-6 Astra, that keep working between conversations using their own cloud computer and the apps you connect. Your first dot is included in Pro and Business Premium at no extra cost, with an allowance for deeper work. Pro access excludes the EEA, Switzerland, and the UK.
- [OpenAI Shares a PHP Version Checker and Codex Skills](https://aicatchup.com/news/openai-php-version-checker-codex-skill): OpenAI has published a PHP Version Checker repository with Codex skills for identifying PHP versions that may need security updates. The checker reads the version a website advertises in its headers or a local PHP installation reports and compares it with the upstream fixed releases from the October 21, 2026 PHP advisory. It is a triage aid, not proof that a system is patched or secure.
- [OpenAI's Decisions API Is Now Generally Available: Fast, Typed Answers from Text, Images and JSON](https://aicatchup.com/news/openai-decisions-api-public-beta): OpenAI's Decisions API is a generally available endpoint for asking structured yes/no, classification, and scoring questions about text, images, or JSON. It left public beta on October 22, 2026, still uses GPT-6 Luna, and OpenAI says it returns typed answers about 10x faster than the Responses API. Uncached input costs $0.10 per million tokens; output tokens are not charged.
- [Claude Security Adds Continuous Scanning for Enterprise Repositories](https://aicatchup.com/news/claude-security-continuous-scanning-enterprise): Anthropic has opened continuous scanning in Claude Security as a public beta for Claude Enterprise. After a deep baseline scan, it re-scans repository changes based on risk and complexity, tracks findings as fixes land, and can surface findings through an API and MCP connector.
- [Cursor Adds GLM 5.3 and GLM 5.3 Flash to Its Model Picker](https://aicatchup.com/news/cursor-adds-glm-5-3-models): Cursor lists Z.ai's GLM 5.3 and GLM 5.3 Flash in its model picker. GLM 5.3 is for higher-quality work at $1.4 per million input tokens and $4.4 per million output tokens; Flash is smaller and faster at $0.15 and $0.5. Both support Cursor's agent tools and are third-party models: billed from prepaid credit on The X Subscription, and from the Other Models pool on legacy Cursor plans.
- [Cursor Router: Auto Balance, Auto Intelligence, and Cost Modes](https://aicatchup.com/news/cursor-router-auto-intelligence-balance): Cursor Router is the routing system behind Auto, and Cursor documents it as available only on Teams and Enterprise plans. Open the model picker, select Auto, and pick Cost, Balance, or Intelligence under Optimize For. Cost mode bills as a Cursor-model request from the weekly pool; Balance and Intelligence bill at the list price of whichever model the request is routed to.
- [Gemini 4 Argon Is Now Available in Cursor](https://aicatchup.com/news/cursor-gemini-4-argon-availability): A practical look at the model's placement in Cursor's usage pools and its listed per-token rates.
- [Gemini 4 Argon Is Now Available to Developers and Subscribers: Benchmarks, Pricing, and Where to Use It](https://aicatchup.com/news/gemini-4-argon-launch): Google made Gemini 4 Argon available on October 13, 2026 in the paid Gemini API, the Gemini app for Pro and Ultra, and Antigravity (Ultra, then Pro by October 16). It ranks third on the Vals Index at 69.24%, first if Claude's fallback-answered refusals count as failures, and costs $2 input and $10 output per million tokens through November 30.
- [OpenAI Models and Codex Are Coming to Google Cloud](https://aicatchup.com/news/openai-models-google-cloud-model-garden): OpenAI and Google Cloud plan to make OpenAI frontier models and Codex agentic workflows available in Google Cloud's Model Garden. The announcement says frontier models can be purchased through Google Cloud Marketplace and general availability is planned in the coming weeks, but it does not name the models or a launch date.
- [OpenAI Adds 12-Hour Prompt Caching to the Responses API](https://aicatchup.com/news/openai-12-hour-prompt-caching-api): OpenAI added an optional 12-hour prompt-cache lifetime to the Responses API for GPT-5.6 and later. Developers can keep matching prompt prefixes eligible for cached-input pricing across longer gaps by setting `prompt_cache_options.ttl` to `12h`; the default remains 30 minutes. The longer cache write costs more, so the tradeoff depends on reuse.
- [Anthropic Opens OSS Scanner for Critical Open-Source Projects](https://aicatchup.com/news/anthropic-oss-scanner-open-source-projects): Anthropic's OSS Scanner is a free, opt-in service that gives eligible open-source repositories periodic security scans using its strongest models. Core maintainers apply through a pull request to the scanner's repository; reports are model-generated, sent without human review, and may include candidate patches. Anthropic says projects are selected case by case, and some findings may be false positives.
- [OpenAI Reports Stronger Coding Results in COBOL, Fortran, and Visual Basic](https://aicatchup.com/news/openai-low-resource-programming-languages): OpenAI says its latest models improved on two software-engineering evaluations for low-resource languages. On MultiLangBench, GPT-6 Astra averaged 87.4% at maximum reasoning effort across 20 programming languages; on a COBOL migration test, GPT-6.1 Sol scored 86.6% at low reasoning effort. These are OpenAI-designed evaluations, not a guarantee of production migration quality.
- [Scheduled Claude Agents: Pick the Right Surface, Then Stop Them Failing Silently](https://aicatchup.com/practices/scheduled-claude-agent-automations): Run repository work on Claude Code cloud scheduled tasks, formerly routines, and production automations on Claude Managed Agents scheduled deployments, which add minute-level cron and a dollar cap per run. Whichever surface you pick, apply Anthropic's six rules: bookmarks, unreadable-not-quiet, re-check before posting, confirm then record, read-only preferences, least privilege.
- [Cursor Cloud Agents Add Interactive Canvases and an Environments API](https://aicatchup.com/news/cursor-cloud-agent-canvases-environments-api): Cursor's latest update adds interactive canvases to Cloud Agents, a beta API for creating and managing cloud environments, and a reported 5x faster load for large environments. The canvases can be shared with a team by link, while the Environments API makes environment management programmatic.
- [Cursor Start Is Now a Legacy Plan: What ₹649 India Subscribers Keep and What Replaced It](https://aicatchup.com/news/cursor-start-india-plan): Cursor Start, the ₹649-a-month India plan launched July 28, 2026, is now a legacy plan. New customers choose The X Subscription, where Cursor names Plus ($8 a month before tax) as the comparable tier. Existing subscribers keep Start, its Cloud Agents and its monthly pool until they switch or cancel, and switching is permanent.
- [Anthropic Adds OpenTelemetry Tracing to Claude Platform in Public Beta](https://aicatchup.com/news/anthropic-claude-platform-opentelemetry-tracing): Anthropic's Claude Platform now offers OpenTelemetry tracing in public beta for Messages API requests, Claude Managed Agents sessions, and Claude SDKs. You can send traces to an OTLP-compatible observability platform, choose metadata-only or full-content mode, and inspect model, tool, latency, and token signals while accounting for beta and data-handling limits.

## Popular Topics

- [OpenAI](https://aicatchup.com/topic/openai): 80 articles
- [Anthropic](https://aicatchup.com/topic/anthropic): 79 articles
- [Developer Tools](https://aicatchup.com/topic/developer-tools): 73 articles
- [2026 Trends](https://aicatchup.com/topic/2026-trends): 71 articles
- [AI Coding Tools](https://aicatchup.com/topic/ai-coding-tools): 53 articles
- [Claude](https://aicatchup.com/topic/claude): 53 articles
- [Claude Code](https://aicatchup.com/topic/claude-code): 52 articles
- [Codex](https://aicatchup.com/topic/codex): 50 articles
- [Cursor](https://aicatchup.com/topic/cursor): 41 articles
- [ChatGPT](https://aicatchup.com/topic/chatgpt): 34 articles
- [AI Agents](https://aicatchup.com/topic/ai-agents): 25 articles
- [Agents](https://aicatchup.com/topic/agents): 23 articles
- [Pricing](https://aicatchup.com/topic/pricing): 18 articles
- [Workflow](https://aicatchup.com/topic/workflow): 18 articles
- [MCP](https://aicatchup.com/topic/mcp): 15 articles
- [Agentic Coding](https://aicatchup.com/topic/agentic-coding): 13 articles
- [Security](https://aicatchup.com/topic/security): 13 articles
- [Computer Use](https://aicatchup.com/topic/computer-use): 12 articles
- [Model Launch](https://aicatchup.com/topic/model-launch): 12 articles
- [Plugins](https://aicatchup.com/topic/plugins): 12 articles
- [GPT-5.6](https://aicatchup.com/topic/gpt-5-6): 11 articles
- [API](https://aicatchup.com/topic/api): 10 articles
- [Best Practices](https://aicatchup.com/topic/best-practices): 10 articles
- [Enterprise](https://aicatchup.com/topic/enterprise): 10 articles

## Machine-Readable Resources

- [Complete content catalog](https://aicatchup.com/llms-full.txt): Every article, newsletter edition, and picks page with metadata
- [RSS feed](https://aicatchup.com/rss.xml): Latest articles and newsletter editions
- [XML sitemap](https://aicatchup.com/sitemap.xml): Canonical indexable URLs with freshness dates

## Publisher

- [About AI Catchup](https://aicatchup.com/about)
- [Editorial Policy](https://aicatchup.com/editorial-policy)
- [Contact](https://aicatchup.com/contact)