In partnership with

n

MiniMax just dropped H3, and it's not playing around.

This open-weight model generates 2K video with native stereo audio, and you can control it with text, images, video, or sound.

The Chinese AI lab is handing everyone the keys to professional-grade video generation.

But wait, there's more:

  • OpenAI is fighting Apple's lawsuit by leaking their own emails

  • Four Chinese open-weight models are matching GPT-4o at a fraction of the cost

I'm Alex. Welcome to L8R by Innov8.

Let's dive deep 🐰

In today's post:

• MiniMax drops H3: 2K video generation goes open-weight
• OpenAI leaks receipts to fight Apple's trade secret lawsuit
• Chinese open models match GPT-4o at 1% of the cost

Forget Sora: MiniMax Drops Open 2K Video Model

Chinese AI startup MiniMax just dropped a massive bomb on the video AI market with H3.

This new open-weight model generates stunning 2K video and native stereo audio all at once.

It gives builders the exact tools they need to bypass closed systems like OpenAI's Sora.

The details:
  • H3 pumps out 2K resolution video with perfectly synced stereo audio built right in.

  • You can prompt the model using text, images, existing video clips, or even audio files.

  • MiniMax released H3 as an open-weight model, meaning developers can download the weights and build their own apps.

  • It handles both creating videos from scratch and editing existing footage like a pro.

Why it matters:
  • Most top-tier video models keep you locked behind a web interface or a paid API.

  • MiniMax is giving away the actual weights for H3 so you can run and tweak it yourself.

  • Generating video and audio together also fixes a huge headache for creators who usually have to match sound to visuals later.

  • This puts serious pressure on closed models from OpenAI and Runway.

💡 L8R's Take:
  • Open-weight video is the future, and MiniMax knows it.

  • While Sora hides behind closed doors, H3 lets real builders get to work right now.

  • If you want to win the AI video race, you have to let the community play with your toys.

From Our Partner

Turn AI into Your Income Engine

Ready to transform artificial intelligence from a buzzword into your personal revenue generator

HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.

Inside you'll discover:

  • A curated collection of 200+ profitable opportunities spanning content creation, e-commerce, gaming, and emerging digital markets—each vetted for real-world potential

  • Step-by-step implementation guides designed for beginners, making AI accessible regardless of your technical background

  • Cutting-edge strategies aligned with current market trends, ensuring your ventures stay ahead of the curve

Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.

OpenAI Claps Back at Apple with Leaked Texts

OpenAI is playing hardball.

The AI giant just leaked its own internal emails and iMessages to fight a massive lawsuit from Apple.

Apple claims OpenAI stole secret AI hardware designs, but OpenAI says Apple is just scared of losing the hardware race.

The details:
  • Apple sued OpenAI, claiming former Apple workers stole confidential AI hardware secrets when they changed jobs.

  • OpenAI leaked internal recruiter messages and onboarding logs to prove they did not steal anything.

  • OpenAI calls the lawsuit an aggressive trick to stop competition in the new AI chip market.

  • This fight shows how desperate Big Tech companies are to control the future of custom AI hardware.

Why it matters:
  • The fight for AI talent is turning ugly.

  • Big Tech companies are using lawsuits to trap their best workers and slow down rivals. Custom AI chips are the next gold rush.

  • If Apple wins, it will be much harder for AI labs to hire the hardware experts they need.

💡 L8R's Take:
  • Apple is just throwing a tantrum because they are losing their top talent.

  • This public fight is exactly the kind of transparency we need in tech.

From Our Partners

Save 40% on 1000+ AI APIs

AI costs don't usually explode overnight. They grow quietly through duplicate requests, expensive routing and poor visibility.

Mesh API helps engineering teams spot the waste before finance does.

A global e-commerce company was able to reduce spend by 78%.

Free Chinese AI Models Crush OpenAI's Prices

Developers are ditching expensive AI models for cheap Chinese alternatives.

High-performance open-weight models are now matching GPT-4o and Claude 3.5 Sonnet in heavy coding tasks.

You can run these beasts yourself without paying crazy API fees.

The details:
  • Alibaba's Qwen 2 (72B) and DeepSeek-V2.5 are leading this massive shift.
    • Reddit users report these models perform just as well as Claude 3.5 Sonnet for complex web development.

  • DeepSeek-V2.5 uses a highly efficient Mixture-of-Experts setup to keep running costs incredibly low.

  • You can download the weights directly from Hugging Face to host locally or use cheap APIs like OpenRouter.

Why it matters:
  • The gap between closed US models and open international models is completely gone.

  • Small teams and solo builders no longer need a massive budget to put top-tier AI into their apps.

  • Running these models locally also means you keep your data private and stay safe from sudden API outages.

💡 L8R's Take:
  • OpenAI and Anthropic should be terrified right now.

  • If I can get GPT-4o level coding for pennies on the dollar, I am never paying premium API prices again.

  • The future of AI is cheap, open, and running on your own hardware..

🚀 Quick L8R Summary

  • MiniMax H3: Chinese startup drops open-weight model that creates 2K video with synced stereo audio from text, images, or clips.

  • OpenAI vs Apple: OpenAI leaks internal messages to fight Apple's lawsuit claiming ex-employees stole hardware secrets.

  • Chinese models: Four open-weight Chinese models now match GPT-4o and Claude 3.5 Sonnet on coding tasks at a fraction of the cost.

📩 Innathe L8R engane undarunnu 👇?

We read every reply - just reply to this email and let us know how we can improve !

Appo adutha L8R il kanaam bie…👋