Good Morning! Here's what I have for you in today's newsletter:

  • Google adds XR Blocks to AI Studio for vibe coding 3D apps

  • Anthropic raises Claude API requests to 128 MB, four times the old cap

  • xAI lets developers set custom moderation rules in the Grok API

  • Remove your exposed personal info from the internet with help from Claude

  • 4 new AI tools worth trying today

AI TOOLS

Google added XR Blocks support to AI Studio, letting users build interactive 3D apps from a prompt. They run in Chrome on Android XR headsets, with a desktop Chrome simulator for testing without the hardware.

  • Apps can track hands and respond to pinch gestures, letting users grab virtual objects or press buttons without a controller.

  • Objects can interact with real tables and walls, landing on surfaces or bouncing off the room around you.

  • Text descriptions and photos can become 360° 3D holograms. XR Blocks, the library behind these features, is open source on GitHub.

An XR idea is easier to judge when you can move things around rather than describe them on paper. AI Studio brings prompting, hand controls, and room-aware physics into the same prototyping workflow. The desktop simulator lowers the hardware barrier, though a headset is still where you’ll see how the experience feels in a real room.

TOGETHER WITH LOVABLE

Test that business idea this weekend

  • Skip the funding round and co-founder hunt

  • Describe the business, get the real product

  • Take payments from real customers immediately

  • Find out if your idea works in days

  • Keep going, or try the next one

Discount of $5 applies to first month of any paid plan. You need to sign up via this webpage in order to claim the discount. The discount applies to new and returning users without an active paid plan. Terms apply.

AI DEVELOPERS

Anthropic raised the request size limit for its Messages and Token Counting APIs from 32 MB to 128 MB. The change addresses image-heavy jobs that hit the upload ceiling before filling the model’s context window.

  • Requests containing more than 20 images now support up to 3,000 pixels per image, up from 2,000, preserving more detail in screenshots.

  • The larger limit applies only to Anthropic’s direct API. Cloud platforms retain caps of 20 MB to 32 MB, with Amazon Bedrock at the lower end.

  • Token limits and context windows haven’t changed. Neither has pricing, and developers don’t need a new SDK version or beta header.

A request can be too large to send even when the model has room to read it. This update gives screenshot-heavy sessions and large document batches more breathing room before developers need to split them up. Apps using multiple providers still need route-specific size checks, since the higher cap doesn’t carry over to cloud platforms.

AI SAFETY

xAI added a guidelines option to Grok API text requests, giving developers control over how responses handle certain content categories. Grok applies the instructions while composing its answer, and all API customers can use them.

  • Controls cover sexual content, violence, and profanity. Educational and clinical questions remain supported, as do history and news responses without graphic detail.

  • Unspecified categories keep the default settings. Guidelines apply to individual requests, so an app must include them on every call to maintain them throughout a conversation.

  • The option works with Chat Completions and the Responses API. Multi-agent models and deferred completions aren’t supported.

A classroom assistant and a customer support bot may need different boundaries for their answers. These controls let developers adjust three categories without changing the response format. They also add something to track: leave the guidelines out of a later request, and Grok falls back to its defaults.

HOW TO AI

Find where your details appear and let Claude help you take them down.

Claude in Chrome can help you work through removal forms and change settings that make your accounts easy to find, starting with your phone number. You’ll need a paid Claude plan, desktop Chrome, and the Claude desktop app.

Step 1: Set up Claude in Chrome

Install Anthropic’s official Claude extension and sign in. In the desktop app, connect Chrome and optionally Gmail for removal confirmations. Review the permission mode, then sign in to the social accounts you want to check. Keep final submissions and verification steps in your hands.

Step 2: Request removal

Ask Claude for a checklist of relevant caller ID services, social settings, and official opt-out pages. Search your number on Google and add confirmed listings. Have Claude prepare the Truecaller and Sync.me forms and check Getcontact’s privacy process. Complete CAPTCHAs and verification yourself. Record submission dates, confirmations, and anything still waiting on you.

Step 3: Change your privacy settings

Have Claude review Facebook’s phone-number discovery options and X’s “Let people who have your phone number find you” setting. Ask it to show the choices before saving. Review each account separately. For phone-only controls in apps such as Zalo and Telegram, get the instructions and make the changes on your phone.

Step 4: Check again each week

Schedule a Sunday review with an explicit timezone. Ask Claude to check confirmation emails, review the settings you changed, and flag unresolved requests after 30 days. Have it propose follow-ups for your approval rather than resend automatically.

Keep the computer and apps available for browser-based checks. Search your number again after removal and record any returning listings. Removing information from one service doesn’t erase every copy, so work through the checklist over time.

P.S. You can access all my AI courses, trainings, prompts library, and $500+ worth of AI tool discounts here.

Anthropic renamed its Max plans. Max 5x becomes Max at $100 a month, while Max 20x becomes Ultra at $200. Verified Ultra accounts no longer have the five-hour limit, but weekly limits remain. Newer accounts keep a temporary cap until verification.

Google opened an Antigravity plugins marketplace with entries from Google, Atlassian, Figma, Stripe, and others. Plugins can combine skills, MCP servers, rules, hooks, and custom agents. They’re listed in the app’s Customizations panel.

Nous Research released Memory Wiki, an official Hermes plugin that organizes local chat history into topic pages and daily logs. It also shows what Hermes remembers about you, without making AI calls or rewriting the original conversations.

🔍 Brave: Says Ask Brave now answers simple questions about 70% faster and reads ten times as many search results when preparing an answer.

🚨 OpenShip: Launched Agents that investigate failed deployments and service incidents in your cloud, then apply fixes after approval using Claude Code or Codex.

🌐 Firecrawl: Added support for Vercel’s open-source eve framework, bringing agents web search, scraping, crawling, and 100+ data providers through Alexandria.

🎙️ Riverside: Launched Agents for recording preparation, episode editing, social clips, and post scheduling. Publishing still requires approval.

Which image is real?

Login or Subscribe to participate

THAT’S IT FOR TODAY

Thanks for making it to the end! I put my heart into every email I send, I hope you are enjoying it. Let me know your thoughts so I can make the next one even better!

See you tomorrow :)

- Dr. Alvaro Cintas

✓ Full archive of premium guides with ready-to-use prompts

✓ Structured AI courses (step-by-step, start-to-finish)

✓ Every upcoming premium tutorial

✓ $500+ worth of AI tools