Welcome back to Visually AI!
Today’s reading time is about 5 minutes.
In today’s edition:
• 2025: A Year of Personal Growth & Gratitude
• The AI Launches That Changed Everything
• What’s Coming in 2026
• Image & Video Prompts
2025 wasn’t just a year of incremental improvements for AI, and and we will get into that shortly.
On a personal note, this year has been transformative:
• Creative Director at Prime Video: Leading creative on projects that are redefining what’s possible with generative AI
• AI VFX for House of David: Bringing seasons 1 & 2 to life with generative workflows
• AI Educator & Consultant: Training teams and sharing knowledge with all of you
And most importantly, this community. We’ve grown to over 11,000 subscribers on Substack and 52,800 followers on Twitter/X.
Thank you for being here.
Lets take a look at what happened in 2025:
Runway Gen-4.5: State-of-the-art motion quality and visual fidelity.
Google Veo 3.1: Native audio generation, multi-shot sequencing.
Kling 2.6: Advanced motion control, native audio synthesis, faster generation.
LTX-2: 20 seconds of continuous video. Finally - longer form content!
Luma Ray3: First model with native HDR support.
Sora 2: Improved motion quality and visual fidelity. Community-driven developments.
PixVerse V5: Advanced visual effects and motion control.
Hailuo 02: High-quality video with improved consistency.
Seedance 1.0: Multi-shot video generation from text and image.
Wan 2.6: Multimodal video model (text, image, audio). 1080p output, native audio, multilingual.
Wan 2.2: Open-source model. Consumer GPU support (8.19 GB VRAM).
Nano Banana Pro: Reasoning-based image generation. Multi-reference support. You can actually talk to it like an LLM.
FLUX.2: Improved prompt adherence, text rendering, realism.
Adobe Firefly Image 5: Better precision, prompt adherence, realistic images. Creative Suite integration.
Reve Image: Came out of nowhere in March 2025 and instantly topped the leaderboards.
ChatGPT Images: New and improved version (December 2025). Tighter integration with ChatGPT.
Leonardo AI: Continued evolution with improved style control and speed.
Ideogram 3.0: Best-in-class text rendering. Typography that actually works!
Midjourney v7: Continued refinements and style consistency improvements. Still my personal favorite.
Imagen 4 Ultra: High-fidelity image generation with advanced editing capabilities.
Freepik Spaces: Preset workflows and collaborative canvas for node-based creation.
Krea Nodes: Nodal workflows for automated tasks (product placement, multi-angle generation).
GPT-5.2: Latest flagship model with enhanced context handling, real-time routing, and advanced reasoning. Integrated into Microsoft Copilot.
Claude Opus 4.5: Released December 2025. Enterprise accuracy and reasoning.
Gemini 3 Pro: Surprise leader for coding and complex reasoning.
Manus 1.6: Multi-modal creative AI with deep research, visual generation, slide/PowerPoint creation and web development.
DeepSeek R1: Inference-time scaling. Open-source innovation.
Qwen 3: 235B-A22B. Advanced reasoning and multilingual support.
Llama 4: Scout/Maverick variants. Open-source powerhouse.
Grok 3: Released February 2025. Real-time reasoning and web integration.
Video models like Kling 2.6 and Veo 3.1 gave us unprecedented control over motion, lighting, and character consistency.
Image models like Nano Banana Pro started reasoning about images instead of just generating from prompts and changed the entire game.
And LLMs matured into specialized tools. GPT-5.2, Claude Opus 4.5, and Gemini 3 Pro each excelling at different tasks. It’s not about which one is best overall, but which one is best at a specific task.
Shoutout to three of my personal top favorites of 2025!
Image: Midjourney | Midjourney will always hold a place in my heart for it’s effect on the AI community and the uniqueness of it’s images. I don’t think anyone expected a Nano Banana Pro impact coming quite like it did and without doubt it has the most use cases; but Midjourney remains my personal favorite.
Video: Kling | Pound for pound Kling has always been near the top. With each tool gravitating towards different strengths, Kling will still be the first tab I open to test new ideas or prompts; especially after 2.6.
LLM: Manus | A lot less known… but wow. Deep research capabilities, seamless integration with Nano Banana Pro and Veo models, slide and PowerPoint generation, web development, video editing; it's like having an entire creative studio in one tool. The speed is insanely fast, and it actually understands the full creative workflow from concept to execution.
I’ve had some amazing opportunities to work on a variety of projects recently, and I don’t share everything in this newsletter or online. And you know I love testing tools to share.
Take a look at my portfolio to get a quick glimpse of my work:
A quick question for moving forward into 2026…
Are these newsletters actually helping you create better stuff?
Because if not, I want to know what would:
Prompt: epic landscape painting, the sun rises over a dramatic, snow-covered mountain range, casting a brilliant golden and pink light across the clouds and peaks. A single, tiny eagle soars in the vast sky. The air looks crisp and cold. A sense of hope, renewal, and immense scale. Hyper-detailed, beautiful, style of a romanticism painting.
Prompt: cinematic film still, close-up on an astronaut’s helmet, their face is illuminated by complex patterns of light being emitted from a glowing, crystal flower they are holding, reflection of the light patterns in the helmet’s visor, sense of wonder and fear
Prompt: A dynamic tracking shot following a peregrine falcon diving through a canyon at dawn. Camera begins wide, then accelerates into a tight follow, maintaining 3 feet behind the bird as it banks left around rock formations. Morning light creates rim lighting on the falcon’s wings. Depth of field shifts from f/2.8 to f/5.6 mid-shot. Volumetric atmosphere, motion blur on background cliffs, 120fps slow-motion capture feel, ARRI Alexa color science.
Kling 2.6
Prompt: 8-second epic, photorealistic cosmic cinematic scene in deep space (9:16). A colossal holographic clock face floats among distant galaxies and colorful nebulae, formed from swirling starlight. The clock hands sweep toward midnight. Camera: slow constant dolly-in. As the second hand clicks onto 12, the clock erupts into a silent, beautiful supernova of vibrant color and light. From the expanding stardust, the glowing text ‘2026’ appears, formed as a new constellation of stars. Awe-inspiring grand scale, ultra-detailed, luminous, high contrast, clean composition.
Veo 3.1
Thank you for reading.
I hope you have a creative week and a prosperous 2026!
Heather Cooper
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.