Caix

Original vertical 3-screen video with Veo 3.1 — straight out, no editing needed 🔽 #Veo3 #aivideo

Gotta say, the new Veo 3.1 isn’t a big leap — but damn, the visual aesthetics and overall vibe feel way more refined...🤙 Veo 3.1 (left) | Veo 3 (right) #Veo3.1 #flow #aivideo

Veo 3 updates: 1080p HD resolution, 9:16 vertical video support and ~50% lower pricing Veo 3: $0.40/sec (from $0.75/sec) Veo 3 Fast: $0.15/sec (from $0.40/sec)

New in Gemini Canvas: Select and Ask You can now visually edit any part of your web app just by clicking an element and describing the change. No code necessary—just click, describe, and see your updates instantly in preview mode.

🍌 nano banana is here → gemini-2.5-flash-image-preview - SOTA image generation and editing - incredible character consistency - lightning fast available in preview in AI Studio and the Gemini API

@frontrunpro Take it. hahahaha🤣🤣🤣🤣🤣🤣🤣 👇(Caption text) “Why are you kicking me out? I just looked down at my phone for a message...”

Shoutout to whoever cooked this slide. Ru fam = giga goats 🐐🔥

@frontrunpro @ChatGPTapp You nailed it!

hey @ChatGPTapp please upgrade your image model

I tested uploading a main subject image — it handled it well and kept the subject consistent, and nails the headline & layout too.

I love this toilet paper… but I can’t afford it.🧻💸 Lately, I found that making motion posters with #Veo3 is just as impressive. So I put this idea into words — it understands exactly what you want to say, nails the headline & layout. 🌟#aivideo @GoogleLabs @GeminiApp

AI's mission is to eliminate all the tedious tasks and free human creativity. 👇generated by #Veo3.

@elonmusk 🪐Just some feedback: while Grok Imagine looks great overall, every time I watch videos generated by Imagine, the motion still doesn’t feel quite smooth. There’s a slightly choppy, low-frame-rate vibe, maybe it’s a trade-off for faster generation speeds?

Make with Grok Imagine

Surprisingly good results from #Veo when generating native Chinese voice + expression + lip sync. Just use: ...in Mandarin: "你的中文台词" — and it works shockingly well, with a high success rate. Sometimes, you really don’t need those complicated workflows anymore.

@frontrunpro @heavendex @CryptoOmen @FlippingProfits @Soloxbt @xiaosuannai4 @pedrigavifrenki @0xjumpman Amazing!!

1/ $27 M raised in the $LIGHT presale by @heavendex in less than 11 hours — and Frontrun just traced the biggest KOL bags: 🥇 @CryptoOmen $340 k 🥈 @rektober $100 k 🥉 @FlippingProfits $100 k 4️⃣ @Soloxbt $96 k 5️⃣ @xiaosuannai4 $84 k 6️⃣ @pedrigavifrenki $70 k 7️⃣ @0xjumpman $52 k 8️⃣ @CryptoM_7777 $52 k 9️⃣ @gr3gor14n $50 k 🔟 @spacemanbtw $34 k

6️⃣ A powerful multimodal stack: Text + Image + Video. If you're creating for social media, try this Google AI combo: ✅Gemini – for writing + content planning ✅Gemini Image(Imagen)– for generating visuals ✅Gemin Video(Veo3) – for full video output Example workflow: ✦Start with a product doc ✦#Gemini extracts highlights → writes copy ✦Imagen generates visuals ✦Veo3 creates the promo video All in one window — fast, seamless, scalable. If you're also using Veo3 or other AI tools to create content — let's connect! I’m still testing things out and will continue sharing cool finds.

5️⃣ Not sharp enough? Upscale your video to 4K/60fps. By default, Veo3 exports at 720p. But you can boost quality in post using: → Topaz Labs Video AI ✅ Upscale to 4K ✅ 60fps smooth motion ✅ Noise reduction + detail enhancement Highly recommended for creators who care about quality.

Level up your #veo3 workflow with these hacks. After weeks of hands-on testing, these are the most effective techniques I’ve found for controlling #VEO's output.👇 1️⃣ Use JSON prompts to control video details precisely. After extensive testing, I found that writing prompts in JSON format makes Veo3 outputs much more controllable — especially when dealing with complex scenes or emotions. Why? Because AI understands structured, logical language better. Example: JSON{ "scene_setting": { "location": "Modern café, large floor-to-ceiling windows", "time": "Golden hour, sunlight slanting in" }, "characters": { "main_character": "30yo professional woman, white shirt, black glasses", "emotion": "Focused but slightly tired" }, "camera_work": { "movement": "Slow push-in from medium to close-up", "style": "Cinematic depth of field" }, "audio": { "dialogue": "Woman: 'We need to rethink this proposal'", "background": "Soft café ambience, keyboard typing", "note": "(no subtitles)" }, "timing": { "0-3s": "Establish setting and mood", "3-6s": "Push-in, focused expression", "6-8s": "Dialogue, subtle head movement" } } If your writing is strong, you don't have to use JSON — but for structure-heavy scenes, it really helps.