Midjourney Video disrupts the competition
Plus Hailuo 02 Video & Higgsfield Canvas
Total reading time about 6 minutes.
Welcome to Visually AI!
Midjourney V1 Video
Last week, Midjourney sent out this post below via 𝕏…
After nearly 2 years of dominating AI image generation, Midjourney just unlocked video.
Perhaps they wanted to put themselves alongside Kling & Veo, who in my opinion are two of the top contenders right now; but Midjourney brought something special to the table.
It’s called Midjourney V1 and it brings motion to a model known for its stillness. You can hover over any image you’ve generated in Midjourney and animate it!
It’s difficult to describe the joy of being able to animate your favorite Midjourney images, in the way that Midjourney imagined them. If you are a heavy Midjourney user, be prepared to generate in relaxed mode because you will probably run out of fast hours quickly.
Midjourney’s Explore page has a separate video feed now:
Midjourney V1 is an image-to-motion system, and it works differently depending on the mode you’re using:
🟡 Two Ways to Animate:
1. Auto Mode (Default)
You generate a still image (or drop in one)
Tap “Make Video”
Midjourney creates interpolated motion from that image only
No prompt editing — it’s fast, but can be limited
2. Manual Mode
You can prompt directly for motion
Example: “a swirling nebula forming a galaxy, slow rotation”
This gives you more narrative or kinetic control
The UI is still experimental, but it’s more powerful than Auto
In addition, you can extend using auto or manual for up to 20 seconds:
Here’s some of my favorite Midjourney video generations so far:
If you want to watch a live demo and see 30 different video results. Click the link below ↓
💻 Recently On Visually AI Youtube
Midjourney’s New Video Model Demo & First Results!
MiniMax's new Hailuo 02 Video model
And it’s not just Midjourney that is making moves in Video.
Hailuo AI (Minimax) announced a brand new 02 video model:
→ Sharper motion dynamics
→ Better frame coherence
→ More emotional detail retention
→ 1080p output (up to 6 seconds)
→ 720p output (up to 10 seconds)
I used an input image generated with krea.ai's new Krea 1 image model with this prompt:
"POV shot from underwater, an Olympic swimmer breaking the surface in an explosion of bubbles and sunlight. The waterline distorts the transition between struggle and breath. Nike swim gear visible mid-motion."
Hailuo 02 image and text to video is available for all users - videos are more expensive with 02 (25 - 50 credits per 720p; 80 credits per 1080p).
Here’s a look at some of my generations:
Higgsfield Canvas
Higgsfield also dropped a new feature that could quietly change how you work with video.
Higgsfield Canvas lets you:
→ Drop a product exactly where you want it
→ Resize, position, and anchor it in seconds
→ Apply edits like logo placement, outfit swaps, and facial changes — all with pinpoint control and accuracy
The big shift is that where most tools try to predict what you want, Canvas lets you decide. Just upload your video, brush over the area, and your element is placed. That’s it. One click.
Even better — combine it with Higgsfield Speak (for lipsync/talking avatars) and Camera Movement tools for a fully controlled cinematic ad, built in minutes.
→ This replaces hours of production headaches
→ It gives you total creative precision
→ It works inside Higgsfield’s existing interface
Demo Launch Video:
📸 AI Snapshots
Freepik adds Google Veo 3 and Hailuo 02 video in their AI Suite of models and tools to help you streamline visual content production. Currently, Freepik’s Veo 3 is text to video only, while Hailuo 02 supports text and image to video.
Adobe released its Firefly mobile app on iOS and Android, enabling on-the-go generative design. Users can now generate both stills and video from prompts, remix assets, and sync them with Creative Cloud. Adobe also introduced Firefly Boards, a collaborative workspace supporting AI-powered video workflows.
ElevenLabs launched a personal AI conversational assistant called 11.ai. It’s voice-first and you can integrate it with several MCPs to help you manage daily tasks:
Runway introduced a new Chat Mode that lets you generate Gen-4 images, videos, and References directly in the chat interface. I used it to make a short clip with several city scenes. I like it because it offers additional suggestions if you don't have a detailed creative plan. Here's a quick look at how it works, followed by the final video at the end:
Krea launched its first image model called Krea 1. It has incredible aesthetic quality and supports style references and custom trainings for character and style consistency.
Here are some of my first results:
🚀 My Recent Top AI Tools Picks
Topaz Labs Bloom: Enhance images with this new creative upscaler, perfect for AI-generated images. Upscale up to 8x with the option to generate up to 4 variations. This product is different than Topaz Photo AI.
Emergent AI: AI agent vibe coding platform builds apps in minutes with no code.
Pubmed AI: Retrieve, summarize, and organize medical literature in Pubmed's vast library in seconds.
Eleven V3: ElevenLabs’ most expressive text to speech model, supporting 70+ languages, multi-speaker dialogue, and audio tags such as [excited], [sighs], [laughing], and [whispers].
Topaz Labs Astra: Creative video upscale model powered by Topaz Starlight. Choose Subtle or Creative enhancement to add texture and details while denoising and polishing videos.
Manus: General AI agent that uses real-world tasks through step-by-step replays, generates slides, videos w/ Veo 3, and more. I’ve been using this a LOT!
Workflow86: AI-powered workflow automation from a text prompt for businesses.
🖼️ Image Prompts
Prompt: Tracking side shot of a female runner sprinting through downtown [city name] at dusk, weaving through slow city traffic. Headphones in, hair in motion, city lights flicker past like stars. Reflections blur in shop windows as she moves like a current.
Prompt: Low-angle shot, A fashion photoshoot of a woman in a vintage sky blue tracksuit with stripes down the side, sitting against a white wall with a grainy Fuji Superia look and soft shadow falloff. Realistic skin texture, visible pores, micro-expressions, subtle blemishes
🎥 Video Prompts
Midjourney Image & Video | Prompt: A young woman wearing trendy designer white denim overalls stands in front of a vast flower field. The scene is captured in vibrant, pastel colors through hyperrealistic photography.
Veo 3 | Prompt: An ancient underground forge awakens. The camera pushes forward slowly through a long tunnel of blackened stone walls, with glowing cracks pulsating like veins. Embers drift through the air. As the shot approaches the main chamber, a massive pillar of fire erupts upward in slow motion, casting vivid orange and crimson light across molten metal sculptures. Sparks burst outward in ultra detail as the heat visibly distorts the air. Lens flare and flicker effects accentuate the intensity.
Thank you for reading.
Have a creative week!
