Adobe releases Firefly AI Video Generator Public Beta Edition: Easily convert text and images into short videos

Adobe recently announced that its latest Firefly AI video generator has officially entered the public beta stage, allowing users to convert text and images into short videos. This innovative feature is integrated in Adobe's Creative Cloud, allowing creators to easily generate AI videos in familiar tools. The biggest feature of this video generator is the ability to generate video clips with a resolution of 1080p and a speed of 24 frames per second.

Users can customize all aspects of the video generated through a simple and easy-to-use web interface, including style, lens angle, movement mode, and shooting distance. Adobe also revealed that a faster "conception" model is being developed, which can enable faster creation, although the model has a lower resolution. At the same time, Adobe plans to launch a 4K version to meet the needs of higher quality video creation.

In terms of functionality, users can choose from a variety of basic settings such as aspect ratio (16:9) and frame rate (24FPS), as well as eight different camera motion modes such as zoom, pan and tilt. This series of features greatly improves the flexibility and creativity of video generation. Adobe specifically stressed that in response to legal concerns about AI-generated content, the video model was trained using only licensed material and did not use customer data.

To make this feature more user experience, Adobe has launched two new subscription plans: Firefly Standard, which costs $9.99 per month, offers 20 5-second video generation permissions, as well as 6 minutes of audio translation and unlimited image-based Vector tool for use. Firefly Pro charges $29.99 per month, offers 70 videos and 23 minutes of translation services, supports converting videos to more than 20 languages, and matches lip shapes.

While Adobe hopes to stand out with its Creative Cloud ecosystem, competition in AI video generation is intensifying. Companies such as OpenAI's Sora, Runway Gen-3Alpha, and Google's Veo2 are all exploring similarly. At the same time, companies such as Kling, Vidu and MiniMax from China are also actively developing their own video generation technologies.