Google’s Veo 3 Fast doubles AI video generation speed at 720p quality

Google has launched Veo 3 Fast, a significantly faster version of its AI video generation model that doubles the speed of video production while maintaining 720p quality. The upgrade allows both Gemini Pro and Flow users to create videos more quickly and efficiently, with Gemini Pro subscribers getting three daily video generations and Flow Pro users paying just 20 credits per video—a substantial cost reduction from previous pricing.

What you should know: Veo 3 Fast represents more than just speed improvements—it’s part of Google’s broader strategy to scale AI video infrastructure and accommodate more simultaneous users.
• Gemini Pro users receive three daily video generations as part of their subscription, while Gemini Ultra subscribers get significantly more with their 250-per-month plan.
• The faster processing enables Google to test additional features like picture-to-video with voice prompts, where users can speak commands like “turn this selfie into a soap opera starring supercomputers” instead of typing prompts.
• Josh Woodward, who oversees Gemini and Labs at Google, outlined plans to incorporate better subtitle rendering and smoother playback capabilities.

The big picture: Google is integrating AI tools across its entire product ecosystem rather than launching standalone products every quarter, using features like Veo to attract new customers and increase engagement among existing users.

Business applications: Companies can leverage the faster video generation for practical purposes without traditional filming requirements.
• Organizations can create internal update videos and employee onboarding materials entirely through AI generation.
• The reduced wait times and lower costs make it feasible to experiment with multiple video concepts and iterations.

Current limitations: While Veo 3 Fast offers significant improvements, it maintains certain constraints that users should consider.
• Video resolution caps at 720p, which may not satisfy users seeking higher-quality output.
• Subtitle accuracy remains imperfect, with challenges in interpreting sarcasm, irony, and other nuanced emotional expressions.
• The tool serves more as a prototyping and testing platform rather than a replacement for professional video production.

Why this matters: The speed improvements enable creators to spend less time waiting and more time experimenting with video concepts, functioning like “models or miniatures to visualize a scene before filming it for real.” This approach allows filmmakers to test ideas rapidly before investing in expensive productions with human actors and professional editors.