Runway Act-One

Runway Act-One is an AI-powered character animation tool developed by Runway AI, Inc. It was introduced in October 2024 as part of the company’s Gen-3 Alpha video generation system. The technology was designed to make expressive character animation more accessible by allowing creators to use a simple video recording of an actor’s performance to animate a generated character. Instead of relying entirely on traditional motion-capture equipment, complex facial rigs, and extensive manual animation, users could provide a performance video and voice input to generate a character performance that reflects the actor’s expressions, movements, and delivery.

The central idea behind Act-One is to transfer the expressive qualities of a human performance to a digital character. An actor records themselves speaking or performing a scene using a camera, and the system analyses the performance to generate corresponding facial animation. The resulting character can reproduce important details such as eye direction, facial expressions, subtle movements, speaking rhythm, and emotional delivery. This makes the technology useful for creators who want digital characters to communicate with more natural-looking expressions without having to animate every facial movement manually.

One of Act-One’s defining capabilities is its ability to work with generated characters that have different appearances and proportions from the original actor. A human performance can be used to drive a stylised animated character or a more realistic digital person. This gives creators greater freedom when developing characters for animation, storytelling, and other visual projects. The system is not limited to transferring a performance onto a digital copy of the actor; it is designed to translate the performance into a different character while preserving much of its expressive quality.

Act-One focuses on facial performance rather than conventional full-body motion capture. It interprets the actor’s face and delivery to create an animated character performance. This includes the movement of the eyes, facial expressions, and the timing of speech. Preserving these details is important because a convincing character performance depends on more than matching mouth movements to spoken words. Small changes in expression, eye direction, and timing help communicate emotion and make a character appear more believable.

The tool was designed to simplify a process that traditionally required several production stages. Professional character animation can involve motion-capture equipment, multiple camera recordings, facial tracking, manual rigging, and detailed animation adjustments. These requirements can make character performance expensive and time-consuming, particularly for independent creators and small production teams. Act-One offered an alternative approach in which a basic camera recording could serve as the main performance reference. This made expressive character animation more approachable for creators who did not have access to a specialised motion-capture studio.

Another important capability is its application to both animated and live-action-style characters. Creators can use the technology to develop stylised characters for animated storytelling or generate more realistic-looking performances for cinematic scenes. Its ability to work across different character appearances makes it useful for projects that require a variety of visual identities. For example, a creator could record a performance and use it to animate a fictional character, a cartoon-like figure, or a realistic digital person.

Act-One also opened up possibilities for creating dialogue scenes involving multiple characters. A single actor could perform different roles by recording separate performances for each character. The resulting animations could then be edited together to create a conversation between characters with distinct appearances. This approach can help independent filmmakers, animators, and digital storytellers produce dialogue-driven content without needing a separate performer for every character. It can also make it easier to experiment with scripts, character interactions, and different acting styles during the development of a scene.

The quality of a generated performance depends partly on the reference recording. The actor’s facial movements, camera position, lighting, and clarity of performance can influence how well the system interprets the input. A clear recording gives the model useful information about the actor’s expressions and delivery. Since the system is intended to preserve performance details, the quality and consistency of the source video are important considerations when preparing an animation.

Voice performance is another important part of the workflow. Act-One uses voice and video performance inputs to create a more complete character performance. The voice provides speech and delivery information, while the video supplies visual performance details. This combination helps the generated character reflect the actor’s timing and expressive choices. It also makes the technology relevant to dialogue-based animation, where facial movement and speech need to work together.

Act-One was introduced as part of Runway’s broader effort to make generative video and creative tools more controllable for artists. Its development reflected an interest in combining human performance with generative models to create new forms of digital storytelling. Rather than requiring creators to construct every movement through keyframes or traditional animation techniques, it offered a way to use real acting as the foundation for generated character performances.

The technology also included safety measures intended to reduce misuse. Runway described safeguards designed to detect and block attempts to generate content involving public figures, as well as technical measures intended to verify that users had the right to use voices created through its custom voice features. These protections formed part of the company’s wider approach to responsible generative media development. Users should still ensure that they have permission to use any performance recordings, voices, and other materials they submit.

It is important to distinguish Act-One from Runway’s newer character-animation technology. Runway subsequently introduced Act-Two, which continues the idea of using human performance recordings to drive digital characters. The current Act-One research page points users toward Act-Two, which is available across paid Runway plans. Act-Two represents the newer stage of Runway’s character-performance workflow, while Act-One remains the original technology introduced in 2024. Consequently, users researching Act-One should check the current Runway platform for availability rather than assume that the original Gen-3 Alpha interface remains unchanged.

Runway’s wider platform also includes video generation, image generation, editing, audio tools, and creative workflows. These capabilities can complement character animation by helping creators develop scenes, generate visual assets, edit footage, and prepare content for presentation. However, Act-One itself is specifically associated with expressive character performance generation, and it should not be described as a complete video-editing or animation production suite.

Act-One is relevant to several creative fields. Animators can use performance-driven character animation to explore expressions and dialogue scenes. Independent filmmakers can experiment with digital characters and cinematic storytelling. Game and entertainment creators can investigate ways to develop expressive characters for visual content. Educators and content creators can also explore performance-based animation for explanatory or narrative projects. Its main value is the ability to translate a human performance into a generated character performance using a relatively simple video-based workflow.

Overall, Runway Act-One introduced a more accessible approach to expressive digital character animation. Its use of ordinary video and voice performances reduced dependence on specialised motion-capture equipment and made it possible to transfer acting details to characters with different visual appearances. Although Act-Two is now the newer character-performance tool within Runway, Act-One remains an important development in the company’s generative animation history.

Pricing

Runway Act-One does not have a separately published subscription price. It was introduced as a feature within Runway’s Gen-3 Alpha platform. Runway’s current pricing applies to access to its wider AI creative tools, and Act-Two is the newer character-performance option.

Current Runway individual plans include:

• Free – $0/month: 125 one-time credits and 5GB of asset storage.
• Standard – $15/month: 625 monthly credits. Annual billing reduces the equivalent monthly price to $12.
• Pro – $35/month: 2,250 monthly credits. Annual billing reduces the equivalent monthly price to $28.
• Max – $95/month: 9,500 monthly credits. Annual billing reduces the equivalent monthly price to $76.
• Team and Enterprise: Additional workspace and enterprise options are available, with pricing depending on the arrangement.

These are Runway platform prices, not confirmed standalone prices for Act-One. Credit requirements and availability depend on the specific model being used.

Review

Runway Act-One introduced a practical approach to character animation by allowing creators to use ordinary video performances to drive generated characters. Its most notable advantage is its ability to preserve expressive details such as eye movements, facial expressions, and speech delivery. This makes it more useful for performance-driven storytelling than tools that focus only on basic lip synchronization.

The ability to animate characters with different appearances and proportions is another strength. It gives creators flexibility when developing fictional characters and makes it possible to use one actor’s performance as the basis for multiple digital roles. The multi-character dialogue workflow also offers opportunities for independent creators who want to experiment with conversations and narrative scenes without recording separate performances with multiple actors.

However, Act-One should not be treated as a complete character animation or video production application. It focuses on generating character performances, so creators may still need other software for editing, scene composition, sound design, and final production. The quality of the output also depends on the reference recording, and generated results may require further review or editing.

Availability is another consideration because Act-One was introduced as part of Gen-3 Alpha and Runway now directs users toward Act-Two. Its original feature should therefore be understood as an earlier generation of Runway’s character-animation technology rather than assumed to be a separately available current product.

Overall, Runway Act-One was an important step toward making expressive character animation more accessible. Its performance-transfer approach is useful for digital storytelling, animation, and creative video production. The newer Act-Two experience is the more relevant option for users exploring Runway’s current character-performance capabilities.