Runway, a leading artificial intelligence video platform, has upgraded its surprisingly fast Gen-3 Turbo model to include the ability to use input images as the first or last frame of the generated video, as well as a series of Prompt structure examples have been added

Gen-3 Turbo can produce a 10-second video in just over 10 seconds It can only convert from image to video, but when using images as prompts, it provides quality comparable to the slower Gen-3 Alpha model

The first or last frame can be given, and having the AI model animate from or to the image in 5 or 10 second shots is a great way to ensure consistency

The full Gen-3 Alpha model is better than Turbo, with better motion quality and simultaneity, but Turbo is adequate for testing ideas and for quick sequencing of large AI projects

In the latest Gen-3 update, Runway has added example prompts below the text boxes These provide structured prompts based on Runway's guides and include appropriate keywords and motion phrases to improve realism

Each example gives you the best phrasing and space to include customization to get the highest quality motion and realism from the prompt For example, “Handheld tracking shot, follow [subject] at [location]”

If you are using Turbo, these can be used to animate the image used in the prompt

One thing to note is that we have found that using predefined prompt examples leads to more rejections and failed generations than leaving the text field blank and using only images This probably triggered enthusiastic AI moderators

Runway also made Gen-3 Alpha's image-to-video conversion available on its popular iOS app, allowing users to animate any image, including photos taken with cell phones