AI video tools are evolving
From Midjourney Video to Runway Act Two, AI video fits creative workflows
Total reading time around 4 minutes.
Welcome to Visually AI!
AI video is becoming a real creative medium
If you’ve been following my work for a while, you know I’ve tested dozens and dozens of AI video tools right from their inception.
Sometimes they felt like novelty - cool demos with hard to define use cases.
That has started to change as we can start to see realistic ways to use these tools in creative workflows including ad creatives, studio production assets, photography editing, video and VFX editing.
From native audio to character animation to motion stylization, it’s no longer about just generating clips and see what AI spits out. It’s about building full creative pipelines.
Here are some of the latest updates:
📸 Midjourney → Motion
Midjourney’s new video tool transforms stills into 5-second animated clips.
It’s extremely impressive across every style from photorealistic to illustrated images.
The animation preserves emotional tone, aesthetic, and style.
→ Works on all aspect ratios - square, portrait, and landscape
→ Motion direction is handled by the model — no extra prompting needed (but you can add your own prompts)
→ Keeps the Midjourney ‘feel’ while adding a layer of life to previously static art
→ Midjourney just added support for Starting and End Frames with looping capability, PLUS video generation available in Discord now.
This is an example of one of my results using Starting and End Frames in Midjourney with this prompt: Smooth zoom out to show the beautiful young woman with a sad look amongst the cherry blossom blooms
🔊 Veo 3: Lip-Sync + Multi-Platform Access
Google’s Veo 3 quietly became one of the most popular generative tools of 2025.
Now available in Gemini, Flow, and many multi-model platforms, it’s already generated over 40 million videos in seven weeks!
→ Audio is auto-synced and mood-matched (you can’t upload custom yet, but it’s coming)
→ I’ve tested stylized trailers, cinematic POVs, and product visuals
This example is from a single, detailed JSON prompt:
Veo 3’s latest addition: lip sync + audio. It gives you far more control when animating specific characters with natural dialogue and ambient sound.
I generated this Veo lip sync video on Flow, using a Midjourney image upload and this prompt: She sighs and says "It's about time! You can finally lip sync characters in Veo 3!"
🎭 Runway’s Act Two: Motion Capture
Runway launched Act Two, which lets you animate any character with just a reference image and a driving video. It even transfers hand gestures and subtle expressions well.
→ 1-image + 1-video setup (I recorded myself in the app to transfer to a few individual characters)
→ Capable of tracking full body, facial expressions, hand gestures
🌈 Kling KOLORS 2.1
Kling’s new image model update drastically improved prompt control and output quality.
You can now generate fully stylized, animated clips across a range of styles; with batch support for Premium+ users.
Here’s an example generated with Kling Kolors 2.1, animated with Kling 2.1, and audio from Kling:
Prompt - A desert nomad riding a mechanical camel across vast dunes during a sandstorm, their weathered face partially hidden by a tech-enhanced scarf, twin suns setting on the horizon, warm amber and rust color grading with gritty texture
📺 Recently on Visually AI
Google Flow - Veo 3 Results | Portfolio (Heather Cooper)
🚀 My Recent Top AI Tools Picks
Magnific Precision: Now you can creatively enhance/upscale faces without changing the appearance, with additional sliders including Sharpen, Smart Grain, and Ultra Detail.
Compare Character Consistency: Compare Flux Kontext, GPT-Image-1, Runway Gen-4 References, and SeedEdit for character-based images on Replicate.
Composite: Clicks, types, and navigates on your behalf on any website; right in your existing browser.
Kimi K2: Free SOTA model uses a Mixture-of-Experts (MoE) for reasoning and coding tasks.
Weavy: Multi-model platform with creative node-based workflows.
🖼️ Image Prompts
Prompt: A golden retriever dashing through shallow surf at the beach, back angle camera low near waterline, splashes frozen in time, blur trails in waves and paws, afternoon sun glinting off wet fur, overcast day, dramatic clouds
Prompt: An ancient, cloaked archivist figure tends to infinite rows of digital tomes inside a decaying data library. Books flicker between code pages and error logs. Glowing titles include "LOST FUNCTION," "DEPRECATED," and "MEMORY OF V1.0."
🎥 Video Prompts
Image to video prompt: A low-angle shot: A golden retriever joyfully runs through the waves at the beach, splashing water around. The camera gently tracks alongside the dog, capturing the sunlight breaking through the clouds, creating a vibrant, uplifting atmosphere.
Generated with Seedance Pro on Freepik:
Prompt: A chef chops vegetables at a counter, sautés them in a pan, then plates a colorful dish and garnishes it with herbs.
Generated with Hailuo 02 on Freepik:
Thank you for reading.
Have a creative week!
Excited to try Runway's Act Two! Thanks for sharing!
Great work again Heather. Incidentally, can I ask what you use to create your screengrabs please? Hope you had time to chill in Greece although an oxymoron.