Google Finally Unveils Gemini 4 Argon

Welcome to another edition of AI PlanetX.

Google unveils Gemini 4 Argon; Anthropic warns GLM 5.3 lowers cyberattack barriers; DeepSeek targets Nvidia and Huawei chips.

• Hottest AI News

• Top AI & SaaS Tools

• AI Tutorial: Create Realistic AI Voice Clone

• Top AI & Tech News

• AI Art Spotlight

• Resource of the Day: Build Motion Design Studio With Opus 5.5

• Featured AI Video

• Weekly Growth Playbooks to Scale Smarter

Hottest AI News

DeepMind

Google Introduces Gemini 4 Argon With 1M Token Output and Deep Reasoning

Google unveiled Gemini 4 Argon, its latest frontier model for long, complex workflows across coding, finance, legal work and cybersecurity. It is built to reason for longer and handle far larger outputs than earlier Gemini models.

Details:

• Argon lifts the output limit from 64K to 1M tokens. Google reports 77.9% on DeepSWE, 51.3% on AutomationBench and 91.7% on LVBench, plus leading results across finance, legal and tax work

• Google says Argon helped free 300+ TiB of data-center memory and is migrating C/C++ codebases to Rust, including 800K+ Fuchsia lines. It can also find and patch critical vulnerabilities

• Availability begins with trusted cyber defenders. Intro API pricing is $2/M input and $10/M output, cached input is 95% cheaper, and later rates rise to $4/M input and $20/M output

No public release date yet. Paid API users and AI Ultra subscribers are next as Google strengthens safeguards against misuse, prompt injection and misalignment.

Forget SpaceX. These 7 Stocks Could Be Next.

You may not have gotten into SpaceX early, but the space-investing story is far from over.

Record launches, falling costs, government contracts and new technologies are turning space into a commercial market. That growth is creating opportunities across satellites, defense, communications, data and in-space infrastructure.

This free report 7 Best Space Stocks to Own in 2026 reveals seven publicly traded companies positioned across this expanding economy. It explains how each business makes money, what could move its stock and which risks investors should understand.

SpaceX captured Wall Street’s attention. Now see which seven space stocks could benefit from what comes next.

See the 7 Stocks

Anthropic

Anthropic Warns GLM 5.3 Puts Advanced Cyber Exploits Within Easy Reach

Anthropic warns that Z ai’s open-weight GLM-5.3 can autonomously build advanced cyber exploits near Claude Mythos Preview’s level, while weak safeguards make those capabilities much easier for bad actors to use.

Details:

• In ExploitBench, GLM-5.3 built working V8 exploits in 50 of 410 attempts, versus 56 for Claude Mythos Preview. On another exploit test it scored 4%, while older models scored 0%

• GLM-5.3 found unknown browser flaws and chained them into a file-stealing webpage. GLM-5.3-Flash built Chrome exploit for $20.40 with 20 minutes of human help and 8 hours of AI work

• Anthropic says deceptive prompts made GLM-5.3 attempt harmful cyber tasks 64% of the time, rising to 92% with prefilled reasoning and 100% without refusals

NIST says GLM-5.3 is the most cyber-capable open-weight model yet. Anthropic says its public release marks a major threshold for cyberattack risk.

Top AI & SaaS Tools

• AI Emaily (Life-time Deal): AI unified inbox that connects multiple email accounts, filters spam and low-priority messages, and drafts replies in the user’s voice

• America gov: US government chatbot that answers typed or spoken federal-service questions using official sources and removes personal details from uploaded forms [F-R-E-E]

• Team Bots: Shared Grok AI coworkers that learn your team’s context, connect to files and apps, remember info, and help automate ongoing workflows

• Autonomyware: Uses autonomous AI to turn product ideas into manufacturable designs, from concept and CAD to verification and production preparation

• MCP Connectors by Databox: Connect your AI Analyst to business tools to provide context-aware performance explanations and take action

The AI newsletter curated by engineers who build it

TLDR AI is a free daily email curated by Anthropic and ex-Google engineers: the new tools worth trying and the research worth understanding, summarized for 1.1M+ readers.

Subscribe for free

AI Tutorial

How to Build a Realistic AI Voice That Sounds Just Like You

Your voice shouldn't be a bottleneck. With AI voice cloning, create a lifelike digital version that powers your scripts, videos, and voiceovers on demand — no recording sessions required.

1. Visit Fish Audio

   Go to Fish Audio and either create an account or log in to your existing one. Fish Audio provides complimentary credits.

2. Choose Voice Cloning

   From the dashboard, open Create Voice and select Instant Voice Clone. then choose a visibility: Public, Unlisted, or Private — based on whether it’s for a personal project, demo, or professional use.

3. Add project details

   Enter a voice name, short description, and optional tags.

   Tip: Use a descriptive name like “Narration Voice 1” or “Podcast Clone” for clarity.

4. Upload audio sample

   You can either:

   • Upload an existing audio clip (WAV, MP3, or M4A format)

   • Record directly in your browser using your microphone

   Requirements:

   • Minimum 10 seconds of clear speech

   • 30–60 seconds recommended for best quality

   Avoid background noise, music, or overlapping voices — the cleaner the input, the more accurate the cloned tone.

5. Preview before cloning

   Before tapping “Create,” play your sample to check clarity and consistent volume. If it’s muffled or noisy, re-record a short, clean version.

6. Generate cloned voice

   Press “Create” and wait while the system processes your voice.

   This usually takes 20–60 seconds.

   The cloned voice will appear in your Voices list.

7. Test with sample script

   Select your cloned voice and type a test sentence.

   For example:

   “Welcome to my YouTube series, where I teach AI tools.”

   Tap “Generate & Play” to hear your voice clone in action.

8. Use generated audio

   If you’re happy with the results, select “Download” to store the MP3.

   You can then use the file in videos, podcasts, reels, or explainer content.

Note: Create multiple versions with different tones — energetic, calm, formal — to suit different projects.

Top AI & Tech News

• DeepSeek is working to build a hardware-agnostic software layer that makes AI models easier to run on both Nvidia GPUs and Huawei Ascend chips

• HeyGen launched its first all-purpose business video model, built on MiniMax’s H3, for 1¢ per second with audio

• OpenAI and Synopsys are launching GPT-Synopsys, an AI model designed to assist with chip design, verification, and Synopsys EDA workflows

• Hegseth unveiled Project Meridian, a 120-day search for future weapons “from under Earth to beyond the moon,” co-led by Musk, Gingrich and Luckey

AI Art Spotlight

Create Your Own Product Photo Like This Using the Prompt Below:

A world-class luxury magazine cover, 4:5 ratio, featuring [BRAND/PRODUCT] as the cover star. The product floats center frame, lit like a fashion editorial — soft directional studio light with one dramatic color gel in [COLOR] casting a subtle wash. The masthead runs across the top in bold serif or condensed sans-serif typography, styled like Vogue, Time, or Esquire — replace the magazine name with the brand name in oversized letters. A bold cover line dominates the left or bottom third: one punchy headline about the product in large type, followed by 3–4 smaller teaser lines in thin elegant font mimicking real magazine callouts. Issue number, date, barcode, and price tag visible in small print. Background is a clean gradient or seamless studio paper in a tone that complements [COLOR]. The product has hyper-realistic texture and surface detail. Overall feel: if this product had its own magazine, this would be Issue 01. Shot on Hasselblad, photorealistic, print-ready quality.

Model: ChatGPT Images 2.5 / Nano Banana 2

Resource of the Day

The Opus 5.5 Motion Design Studio Playbook

This full course (login to X to view it) shows how to turn Opus 5.5 from a one-prompt video experiment into a repeatable motion design studio. Learn to build a code-to-video pipeline, direct visuals with references and beat-based shot lists, animate UI and type, generate and sync sound, render multiple formats, and run critique loops to fix weak spots.

Featured AI Video

9 Free AI Agent Skills You NEED to Install

Want Weekly Growth Playbooks to Scale Smarter?

Want to find growth trends early? Join weekly playbooks to improve ads, content, tools, and growth decisions with practical steps.

Unlock $10,000+ Worth of F-R-E-E Growth Blueprints, including: 👇