Adobe announced on Wednesday that it will launch a new generative AI-driven video creation and editing tool in a limited beta release later this year, aiming to enhance its offerings for creative professionals.

Named the Adobe Firefly Video Model, this AI tool will complement the existing Firefly suite, which includes applications for generating still images, designs, and vector graphics.

This development positions Adobe to compete in the expanding market for AI video generation tools, which is already being explored by companies like OpenAI with Sora, Stability AI with Stable Video Diffusion, and various startups.

According to Alexandru Costin, Adobe's vice president of generative AI, the tool can create a five-second video clip based on a single prompt and can process both text and image inputs.

Users will have the ability to define camera angles, panning, motion, and zoom.

Costin emphasized that significant investment has been made to ensure the model meets the quality and prompt comprehension expected by videographers, stating that it respects their guidance more effectively than other AI video models.

Adobe clarified that the video model is trained exclusively on public domain or licensed content for which it has permission, avoiding any customer content.

"We only train them on the Adobe Stock database, which includes 400 million curated images, illustrations, and videos that do not contain intellectual property, trademarks, or recognizable characters," Costin explained.

Additionally, Adobe is introducing Generative Extend, a feature within its Premiere Pro video editing software that can extend existing clips by two seconds by generating suitable inserts to fill any gaps.

Costin noted that the initial preview of this tool in April received an overwhelmingly positive response from customers.