Vortex Next Gen Trends

Official Prompting Guide for Sora 2 by OpenAI Structure: — General Scene and Description: • Describe the scene in simple language; • Specify characters, costumes, set design, weather, and other details; • Be as detailed as necessary to create a video that matches your vision. — Cinematography: • Shot type and camera angle: specify, for example, “wide shot, eye level” or “close-up, slight tilt from behind”; • Mood: define the overall tone, e.g., “cinematic and tense,” “playful with tension,” “luxurious anticipation”; • Lens and filtering: you can specify the type of lens and filters, e.g., “32mm / 50mm spherical lenses, light CPL filter”; • Lighting and palette: describe the quality of light and main colors, e.g., “soft daylight from a window with warm fill light from a lamp and a cool reflection from the hallway.” — Actions: • List clear, specific actions or gestures as bullet points; • Try to describe actions as distinct moments or beats tied to time. — Dialogue: • If there’s dialogue in the shot, include short, natural lines; • Keep them concise so the timing fits the clip’s duration. — Background Sounds: • Describe ambient sounds that help set the rhythm or atmosphere; • For example, “hum of coffee machines and murmuring voices” or “rustling paper and footsteps.” https://cookbook.openai.com/examples/sora/sora2_prompting_guide

SORA2 - on Vortex! The all-new Sora 2 Pro brings next-level realism and full cinematic control to your AI videos. Ultra-detailed motion, synced audio, and seamless camera moves turn every generation into a mini movie. Available now exclusively inside Vortex for pro creators. No limits, no watermarks! Create now https://vortex.channel/create.php

They showed AgentKit — remember the Agent inside ChatGPT? Now you can plug that thing into your own app and build any kind of agent — it can click things, or just work through MCP / API, etc. You can see all the tools on the second screenshot; evals are also built in.

Leaderboards. We read: Kling 2.5 Turbo beat everyone, including Veo3. Well, okay, the model is indeed good. But then we look further down: • Minimax Standard strongly outperforms Minimax Pro • Lumalabs Ray3, for some reason, takes 3rd place and beats Veo3 Fast • Seedance is hanging at the bottom Am I the only one confused here? Who exactly are these leaderboards being tested on?

Kling is the best at gymnastics. Trained like a rex. The others are funny. Minimah is in second place.

Google has rolled out another interesting thing — Learn Your Way. The idea: you take a boring chapter from a dull textbook, feed it to the neural nets, and in return you get the same chapter, but as if it were made specifically for you and your interests. Basketball fan? Newton’s laws are explained through the ball and the hoop. Love art? Economics turns into art auctions. And it’s not just about swapping in new examples. It generates different formats: a mind map for visual learners, audio lessons in a “teacher–student” style, interactive timelines, tests that adapt to your mistakes, and so on.

An interesting use of Nanabanana – the very same infinite zoom. Notably, it’s all packaged into an app on AI Google Studio, where you can upload your own image and create this zoom over its parts by simply selecting the desired area. Careful, it’s highly addictive – you can spend a long time watching where it all leads. Try it here (the generation isn’t very fast, the video is heavily sped up): https://aistudio.google.com/apps/bundled/enhance?showPreview=true&showAssistant=true

JetPave AI - I Helped My Friend Fix His PCB With a Secret AI Tool My friend Romain was struggling with endless PCB errors until I showed him a secret AI tool called JetPave. In this video, I reveal how JetPave takes you from idea to factory-ready design with requirements analysis, function definition, solution selection, and schematics you can download as a .sch file and open in KiCad. It even checks market need, picks components, and eliminates costly human mistakes. I put it to the test by designing a smart cat feeder, and the results were insane. Go to https://www.jetpave.com/ to start! https://www.youtube.com/watch?v=mxUI8kjXfFA

Kling 2.5 Update 1. Smarter prompt following & better timing The model now handles complex, multi-step instructions with stronger temporal logic, enabling users to create richer stories, scene transitions, and character interactions. Static images can be turned into dynamic videos with coherent flow. 2. Smoother, more stable motion With reinforcement learning and improved training data, the model produces natural, high-energy character and camera movements while avoiding glitches or distortions in complex scenes. 3. Style consistency Advanced conditioning ensures every frame matches the reference image’s colors, lighting, and atmosphere, even in fast, dynamic videos. 4. Lower cost, higher value 5 seconds of 1080p now cost 25 credits (down from 35). That’s 1,000+ videos/month on Ultra and 320 on Premierusing 2.5 Turbo.

Suno Version 5 soon All that’s known — sometime in about two weeks, at most a month. Hopefully, it will be just as cool as the previous updates.

Alibaba has released the video model Wan 2.2 14B Animate for transferring animation to characters or removing them from a frame. It captures subtle movements of facial expressions and even fingers. There are two approaches: • Animation mode — generates an animated video with a character based on a provided image and a reference video with the desired movement. • Replacement mode — removes a character from the frame and replaces them with the one from the image, seamlessly blending them into the environment with recalculated lighting.

Luma Labs has released a new video model – Ray 3. The new model can generate in 1080p 16-bit HDR, with a duration of 5 or 9 seconds. ➖ The cherry-picked samples look nice – although it’s still clear that the level of detail isn’t very high, especially in distant shots. Faces of people in the background also get blurry. The marketing team, as usual, likes to sell what doesn’t quite exist yet (the release page makes some rather bold claims about the model’s quality, reasoning, and other capabilities). ➖ There’s a Draft Mode, which generates faster but only 5 seconds and in terrible quality (640 × 352). You can only generate for free in this mode.

Meanwhile, the Higgs team launched Higgsfield Fashion Factory. • You choose a background • Create a character • Generate an initial photo set • Upload clothes • Click “change outfit” • Get a full photoshoot in the new clothes The Higgs team is constantly experimenting with packaging their existing features into mini-products. Unfortunately, the Factory can’t be tested for free. https://higgsfield.ai/image/soul?modal-fashion-factory=true

Guys, Adobe has added the ability to edit text on images in its Firefly Boards (beta) 🔥 I tried it out, it works!

Lucy You’ll laugh, but we’ve got a new video generator. And a fast one at that. Interestingly, it comes from the Decart.ai team, who recently dropped Mirage — a real-time generator (or rather, re-skinner) of worlds, and Oasis 2.0 — a Minecraft world generator. Which leads us to think that world generators and video generators will go hand in hand (like Google’s Veo and Genie) and complement each other. The Lucy 14B generator is pretty hefty in terms of parameters (there’s also a 5B version). They claim it can generate a 5-second clip in 6 seconds, but in reality it’s around 12 seconds — which, you’ll agree, is still not bad. Some sources mention clips up to 10 seconds long, but on Fal it seems limited to 5. 720p. https://fal.ai/models/decart/lucy-14b/image-to-video/playground

KeenTools Cloud API – the first fully automated solution for building 3D head models from photos with consistent topology. Classic workflow: 3–5–7 photos as input. Even two can work. After 20, quality doesn’t improve. Doesn’t handle hair yet. Glasses and accessories don’t interfere – it “sees through” them. Also: – Full customization: any topologies, blendshapes, UV maps – Works with regular photos – Automatic detection of camera positions and focal length – Built-in texturing with color correction – Facial expression support and export to blendshapes – Cloud architecture: secure, GDPR-compliant, scalable Applications: cosmetic surgery, game development, VFX, virtual shopping, machine learning, and more. In short, it’s a game-changer – check out the video.

8 codes that will help ChatGPT give more accurate answers for your tasks 🫡 ▪️ ELI5 – an explanation “as if for a five-year-old.” Perfect when you need to simplify complex topics quickly. ▪️ TL;DR – a summary of a long text. Great for research or lengthy articles. ▪️ Jargonize – turns text into a more professional version, suitable for LinkedIn or business content. ▪️ Humanize – adds a human touch to text and makes it more conversational 🤏 ▪️ Step-by-Step – makes the AI explain solutions step by step. ▪️ Compare & Contrast – generates comparisons of different options, highlighting pros and cons. ▪️ Expand – develops a short text into a full paragraph or article, adding details and examples. Perfect for content and posts. ▪️ Roleplay – makes the model take on a specific role: analyst, marketer, psychologist. Helps you get precise, specialized answers. Save this — it will come in handy for everyone 🆗

Higgsfield released UGC Factory based on Nano Banana and Veo 3. What I like about the Higgses is that they constantly test new products on hamsters, they always start from the product, piecing together puzzles from APIs to different generators. They go viral, intrigue, provoke. What I don’t like about the Higgses is that almost nothing is left of their own model — they’re just an ordinary participant in the aggregation of everything that moves. Although Speak 2.0 turned out pretty decent. UGC Factory is about maximizing the simplification of creating ads and not just ads, but TikTok-like videos in general. • You choose a model and a product/object • Generate 4 images with Banana • Add Action • Add Audio • Choose Audio Styles • Add Background That’s it — no visual prompts or prompts the size of a house.

Apple today released FastVLM and MobileCLIP2 on Hugging Face. An 85× speed boost and 3.4× size reduction for real-time image and video processing directly on your device! Real-time video subtitles in the browser — no cloud access required. Impressive. https://huggingface.co/spaces/apple/fastvlm-webgpu

There’s some buzz in the news about a presentation generator from Kimi AI. Although Kimi AI is actually a solid Chinese LLM, it feels a bit odd to see such a sub-product from them. Nobody mentions that the entire product interface is in Chinese — but you can sign up with Google and you’ll get a familiar prompt input window with the letters “PPT” framed by Chinese characters. I asked it to make a presentation about cats: Pros: • Writes fairly meaningful texts • Lets you choose templates • Templates don’t look as awful as those from many other slide generators Cons: • About a quarter of the backgrounds contain Chinese characters • Images are generated stylishly and match the template, but have NOTHING to do with the presentation topic (the random cactuses were especially funny) Conclusion: Not usable. https://www.kimi.com/kimiplus/cvvm7bkheutnihqi2100