Hi Reader,

This week, we’ve crossed a new frontier: the arrival of mainstream AI video creation at your fingertips.From text-to-5-second clips in your Bing app, to upscaling AI content into 4K with a click, the creative workflow has been fully transformed.

Whether you’re a founder, marketer, creator, or just curious, this is a week where the future hit fast-forward.

In today’s breakdown:

• Microsoft unlocks free OpenAI Sora inside Bing
• OpenAI launches o3-pro, their most advanced reasoning model
• Krea 1 , a stunning new image model with pixel-level control
• Midjourney enters video generation and needs your help
• Topaz Labs unveils Astra, a groundbreaking AI video upscaler
• Runway launches Chat for real-time creative teamwork
• ByteDance drops Seedance 1.0 and it’s beating Sora and Veo

Let’s dive in.

Microsoft Brings OpenAI’s Sora to Everyone Free in the Bing App

Yes, you read that right. Microsoft is now offering free access to OpenAI’s Sora video generation model inside the Bing app for iOS and Android. It’s the first major rollout of Sora to the general public, no invite, no waitlist, no fees.

Key details:

• 5-Second AI Videos from Text​
  You can type a short prompt and instantly generate 5-second videos powered by Sora.

• No Subscription Required​
  This is 100% free. No API keys, credits, or premium plans. It’s built into the existing Bing app.

• Mobile-First + Coming to Web​
  Works now on mobile (iOS + Android). A web version is in development, expected later this summer.

• Direct Download and Sharing​
  Once generated, videos can be downloaded or posted straight from your phone, ideal for creators on the go.

This is the most accessible way to try OpenAI’s Sora to date and it’s shockingly good.

Try Bing Video Creator here 👇

​🔗 Download Microsoft Bing on Android​

🔗Download Microsoft Bing on iPhone​

This Week’s Viral Content: Stormtroopers Take Over Veo 3

These short films combine realistic lighting, gritty close-ups, and movie-style camera movement all generated from a text prompt, no crew or camera required. If you've seen those jaw-dropping Star Wars edits floating around on LinkedIn and X, this is how they were made.

Watch my full workflow tutorial to recreate these cinematic Veo 3 videos 👇

​

You asked, I delivered. This Youtube tutorial includes the exact prompt structure I (and many other creators) use to consistently get high-drama, cinematic results from Google Veo 3.
​

Bonus point

Grab the most optimized 10+ Veo 3 Prompt + DETAILED Workflow AND Guide here 👇
​https://creativesclub.store/products/google-veo-3-vlogs-master-prompt-instructions​

This includes prompt formatting, camera guidance, and style tags that elevate your generations from “AI-looking” to Netflix-ready.

And if you want to access to Google Veo 3 without spending a $250/month plan to generate videos, I have the perfect plan for you 👀
​

You can recreate Google Veo3 videos inside of Freepik starting from as little as $17.15 a month.

​
Try Google Veo3 in Freepik here 👉 https://lnkd.in/gs7upqys​

Midjourney Begins Testing Its First Video Model

After dominating AI image generation, Midjourney is quietly shifting gears and entering the AI video space. It’s currently running user tests on its V1 Video Model, and asking community members to rate early results.

Key details:

• Not Public Yet, But Close​
  The model isn’t released for prompts just yet, but hundreds of videos are being shared internally with rating prompts, likely to refine the output.

• Style Retention from MJ Images​
  Early videos maintain that signature Midjourney aesthetic, dreamlike, high-contrast, cinematic.

• User Feedback Driven​
  The devs are asking for community votes to guide the release , which could suggest a summer launch.

Midjourney Video V1 is real. And coming fast. 👇

​Help rate Midjourney videos here

OpenAI Launches o3-pro: Their Smartest, Most Capable Model Yet

While o3 was already powerful, OpenAI just turned the dial further with o3-pro, an upgraded version designed for deep, extended reasoning in fields like math, science, law, and code.

Key details:

• Outperforms o1-pro and o3​
  o3-pro replaces the earlier “pro” model (o1-pro), and shows a 64% win-rate over o3 in expert evaluations across accuracy, clarity, and depth.

• Excels in Thought-Heavy Tasks​
  Unlike faster chat models, o3-pro is slower but more reliable when the problem requires careful analysis, such as solving math proofs or debugging code.

• Beats Claude 4 Opus in Benchmarks​
  According to OpenAI, o3-pro scores higher in comprehensive evaluations than Anthropic’s top-tier Claude 4 Opus.

If you need deep thinking, not just fast talking ,this is the model.

Try o3-pro via ChatGPT Pro 👉 https://chatgpt.com/​

Krea 1: A Photorealistic Image Generator With Pixel-Level Control

Krea just dropped its first flagship model, Krea 1, and it’s already earning praise for producing shockingly real, photorealistic AI images, no signup needed.

Key details:

• 1.5K Native Resolution Output You get beautiful, high-res images right out of the box, with support for upscaling up to 4K.

• Custom Training + Daily Free Generations You can upload your own data and train personalized models for even more control. And yes, it’s free to try daily with no login.

• Precision Control Over Aesthetics From cinematic to editorial to glossy fashion, Krea lets you guide lighting, color, sharpness, and texture.

• Real-Time Canvas Krea’s live canvas lets you adjust and iterate images with brush tools and AI composition in one unified interface.

Try Krea 1 here 👉 https://www.krea.ai/​

Runway Launches Runway Chat: Real-Time Collaboration for Creators

Runway has just added Runway Chat, a new feature designed for collaborative creation letting you and your team prompt, edit, and share feedback in real time, all in one place.

Key details:

• Multi-User Sessions​
  Multiple team members can now jump into a shared workspace to prompt AI tools together, for faster feedback and creative iteration.

• Works Across Modalities​
  Whether you're generating video with Gen-3 Alpha, text, or stills Chat lets you do it as a team.

• Instant Reviews + Revisions​
  Live chat features make it easier to refine prompts and outputs, great for agencies or creators on deadlines.

Try Runway Chat here 👉 https://app.runwayml.com/generate?ask​

Want to learn how to use Runway Chat like a pro?

Check my full workflow breakdown here 👇

Topaz Labs Astra: First-Ever AI Creative Video Upscaler

Topaz Labs, known for making some of the best video enhancement tools on the market just launched Astra, the first upscaler made specifically for AI-generated video content.

Key details:

• Upscale to 4K with Enhanced Detail​
  Astra doesn’t just stretch pixels , it uses AI to reconstruct missing detail and boost clarity in text-to-video clips, AI avatars, or animation renders.

• Two Modes for Creators​
  Choose between “Imaginative” (adds stylized texture) or “Restorative” (boosts realism) perfect for Veo, Sora, or Runway videos.

• Web-Based, No Installation​
  No need to download software. You can drag-and-drop and upscale video directly in-browser.

• Powered by Topaz’s New Starlight AI​
  Built on Topaz’s latest model, Astra represents a leap in frame-level enhancement and stabilization.

Try Astra here👉 https://www.topazlabs.com/topaz-video-ai​

ByteDance Drops Seedance 1.0 And It Beats Veo, Sora, and Kling

ByteDance, the company behind TikTok, just made waves with Seedance 1.0, its new video generation model, and according to benchmark tests, it’s now the top performer across the board.

Key details:

• Outperforms Veo 3, Sora, and Kling 2.0​
  Seedance is now ranked #1 on Artificial Analysis leaderboards ahead of OpenAI’s Sora and Google’s Veo 3.

• 5s Videos in Under a Minute​
  It produces high-quality, 1080p, multi-shot stories with smooth motion, consistent characters, and cinematic transitions.

• Real-Time Camera Steering with APT2​
  A separate research model called Seaweed APT2 lets users steer the camera in real-time, like a game engine.

• Full Integration Coming​
  Seedance will be built into Doubao (ByteDance’s Claude-style chatbot) and video app Jimeng by year-end.

ByteDance is quietly building the Pixar of AI and it just passed the West.

Explore Seedance 1.0 demos here 👇https://seed.bytedance.com/en/​

More wild drops, breakdowns, and tools coming next week.

Catch you next week for another round of breakthroughs.

Stay Creative,

Sebastien Jefferies.

Free: My 100+ AI Toolkit to Supercharge Your Workflow​
​Get your copy here → [Access Now]​