$12 Million Raised to Structure AI Video Production With Preview

Preview has raised $12 million to develop a platform that brings together AI video generation, storyboarding, shot lists, collaboration, traceability, and delivery.

Video generation is only one part of the challenge once a project starts involving multiple shots, versions, models, and collaborators. Preview is positioning itself precisely around this production layer. The company says it has raised $12 million in total funding, led by Sequoia with participation from The General Partnership, to develop an environment for studios and teams working with AI video.

The platform operates more like a production tool than a standalone generator. Projects are organized into stories, scenes, shots, and takes, with a dedicated canvas for each scene. Generated images and videos can then be assigned to shots, reordered, and compared, while a Production area provides Shot List, Kanban, and Storyboard views.

This structure also supports collaboration. Multiple people can work in the same canvas at the same time, follow other team members’ activity, and comment directly on assets. Feedback can remain attached to a take or shot, including through share links used for external review and approval.

Preview also focuses on maintaining recurring elements throughout a production. Characters, locations, and props can be saved as reusable references for future generations. The Camera Bag feature lets teams save combinations of cameras, film stock, lenses, aperture settings, and focal lengths so the same visual direction can be reused across different shots and collaborators.

The platform also extends beyond preproduction. A scene can be exported as an XML file with its assets for continued editing in Final Cut Pro, DaVinci Resolve, or Adobe Premiere. This handoff helps clarify Preview’s positioning: generation remains integrated into the workflow, while the project can still move into existing post-production tools.

Traceability is another part of the product. Preview says it keeps assets, prompts, models used, credits consumed, versions, and comments within the workspace to document how each element of a project was created. Its credit system covers multiple models available through the platform, with a documented conversion rate of 100 credits per dollar.

An agent complements this environment by helping users move from an idea to a script and then to a shot list, suggesting prompts or models, and assisting through revision cycles. The product remains centered on collaboration, however: choices, approvals, and versions are organized around a team rather than a fully delegated autonomous generation process.

According to Stefan Fejes, Preview is already being used on commercial productions for Fortune 100 companies and on several hybrid projects combining traditional filmmaking and AI. He also says 3,000 studios are currently on the waitlist, and that films involving Oscar-winning talent are expected to arrive on streaming platforms later this year. These claims come from the company, and the productions involved have not been publicly named in the sources reviewed.