Google Veo 2 & Imagen 3: What Creators Need to Know About the New AI Models
Key Takeaways
- âś“ Veo 2 generates 4K video up to several minutes long with improved physics and realism
- âś“ Imagen 3 produces photorealistic images across diverse art styles
- âś“ New 'Whisk' tool lets you remix images using subjects, scenes, and styles
- âś“ Available now via VideoFX and ImageFX in Google Labs (waitlist)
What Happened
On December 16, 2024, Google DeepMind announced Veo 2 and an upgraded Imagen 3—their latest video and image generation models that achieve state-of-the-art results in human evaluations.
The new models are available through Google Labs tools VideoFX (for video) and ImageFX (for images), along with a new experimental tool called Whisk that lets users remix images by combining different subjects, scenes, and styles.
“Veo 2 creates incredibly high-quality videos in a wide range of subjects and styles. In head-to-head comparisons judged by human raters, Veo 2 achieved state-of-the-art results against leading models.” — Google Blog
Key Features of Veo 2
Understanding Cinematography: Veo 2 understands film language. Ask for a “low-angle tracking shot” or specify “18mm lens” and it will deliver the appropriate wide-angle aesthetic. Request “shallow depth of field” and it blurs backgrounds appropriately.
Improved Physics & Realism: Unlike earlier AI video models that might “teleport” a basketball into a hoop, Veo 2 renders realistic physics. When a shot misses, you see the actual rebound.
Resolution & Length: Videos can be generated at up to 4K resolution and extended to several minutes in length—a significant improvement over competitors.
Fewer Hallucinations: Google claims Veo 2 produces fewer unwanted artifacts like extra fingers or unexpected objects compared to other models.
Why This Matters for Creators
For YouTube & Social Media Creators
Veo 2’s understanding of cinematography means you can generate B-roll, transitions, and establishing shots that look professionally shot. Specify the exact camera movement and lens style in your prompts to get broadcast-quality footage.
For Marketers & Businesses
The combination of Veo 2’s video capabilities and Imagen 3’s image generation creates a powerful suite for producing marketing content. Generate product visualizations, explainer video clips, and social media assets without expensive production.
For Designers & Artists
The new Whisk tool opens creative possibilities for rapid concept exploration. Upload a subject (your product), a scene (desired environment), and a style reference—Whisk combines them into new variations. Perfect for mood boards, concept art, and creative ideation.
Competition Is Heating Up
Google’s announcement puts pressure on competitors like OpenAI’s Sora, Runway, Pika Labs, and others. For creators, this competition means better tools, faster improvements, and more options.
Try Google's VideoFX
Join the waitlist to access Veo 2 and create state-of-the-art AI videos
Join Waitlist →How to Get Started
VideoFX (for Veo 2 Video Generation)
- Visit labs.google/fx/tools/video-fx
- Sign up for the waitlist
- Once approved, start with simple prompts and iterate
- Use cinematography terms for better results (lens types, shot types, lighting)
ImageFX (for Imagen 3 Images)
- Go to labs.google/fx/tools/image-fx
- Available now in 100+ countries
- Try specific art styles and detailed prompts for best results
Whisk (for Image Remixing)
- Visit labs.google/fx/tools/whisk
- Currently available in the U.S.
- Upload or generate images for subject, scene, and style
- Let the AI combine them into new creations
Safety & Watermarking
All Veo 2 outputs include an invisible SynthID watermark to identify AI-generated content. This helps combat misinformation and ensures transparency about the content’s origin.
Google has restricted generation of public figures and photorealistic likenesses without consent, and has been intentionally measured in rolling out access to manage safety.
The Bottom Line
Veo 2 and Imagen 3 represent a significant leap forward in AI-generated media quality. For creators, this means more powerful tools for ideation, prototyping, and content creation. The key is learning to prompt effectively—using cinematography language for video and detailed style descriptions for images.
While these tools won’t replace professional production for high-stakes content, they’re invaluable for rapid prototyping, social media content, and creative exploration.