Google has widened the rollout of its AI video creation tool, Vids, to more Workspace users, integrating advanced video production capabilities directly within its suite. The tool, initially previewed in April 2024, leverages the power of Google’s Gemini AI to streamline video projects by using document content and custom prompts.
Vids is designed to simplify video production for teams without specialized editing knowledge, with access now extending to accounts such as Business Standard and Enterprise Plus. The rollout is expected to be fully visible within 15 days.
Simplifying Video Creation in the Workplace
The essence of Vids lies in its ability to reduce the complexity of video production. Users can initiate projects by choosing from customizable templates or starting from scratch using the “Help me create” feature.
This AI-driven function suggests scripts and assembles a preliminary storyboard that includes stock media and scene layout recommendations. Customization options allow users to add voiceovers or record content within the app. Vids also supports real-time collaboration, reminiscent of tools like Docs and Slides, enhancing team engagement.
The Vids platform offers features such as automated voiceovers, background image removal, and basic editing functions. These capabilities will be available at no additional cost until December 31, 2025, although Google plans to introduce usage limits starting in 2026, with notifications issued beforehand. Supported browsers include Chrome, Firefox, and Edge on Windows, while others might not fully support all functionalities.
The release of Vids is part of Google’s larger effort to embed AI capabilities into its products. Earlier in May 2024, at Google I/O, the company introduced Veo, which can produce high-resolution, minute-long videos that understand cinematic elements and simulate real-world movement. Demonstrations included collaboration with Donald Glover’s creative studio, Gilga, showcasing Veo’s storytelling potential. Veo has since been incorporated into YouTube Shorts to broaden its appeal for creators.
Adobe’s AI Integration with Firefly Video Model
Weeks prior to this rollout, Adobe unveiled its Firefly Video Model at the Adobe Max conference in October 2024, bringing AI enhancements to Premiere Pro. A standout feature is Generative Extend, which allows users to lengthen clips by up to two seconds and modify video frames for smoother transitions.
Audio adjustments are limited to background ambiance and exclude human voices or music due to rights constraints. Adobe also introduced web-based text-to-video and image-to-video features, limited to five-second, 720p clips. Each video generated through Firefly includes Content Credentials to indicate AI involvement and address copyright transparency concerns.
The Competitive AI Video Market
The introduction of Google Vids comes at a time of significant developments in AI video production, with both top AI startups and Big Tech companies competing to offer innovative solutions. In August, Runway ML launched its Gen-3 Alpha Turbo model, delivering video outputs seven times faster than its predecessor while reducing costs. In September, the company also launched a new API for the model.
Open-source alternatives have also gained momentum, with Pyramid Flow being released in October 2024 by researchers from Peking University and Kuaishou Technology. The model supports up to 10-second clips at 768p resolution and uses pyramidal flow matching, a process that refines video quality in stages, reducing overall processing needs. This technique offers flexibility for developers aiming to modify or integrate the software into custom projects without licensing fees.
Another notable player, MiniMax, backed by Alibaba and Tencent, launched its video-01 model in September 2024. Their model excels at producing high-quality, six-second 720p clips with realistic human motion.
Ethical Concerns and Industry Challenges
The rapid expansion of AI in media creation has not been without its controversies. Adobe, for instance, emphasizes that its Firefly tools are trained on licensed data, ensuring that content is commercially safe. This contrasts with criticisms directed at other models like OpenAI’s unreleased Sora AI video model, which faced scrutiny for potentially using unconsented data during training.
