Contributed by Giridhar Manepalli

We recently released version 1.0 of the MovieLabs Creative Vocabulary to capture this shared vocabulary. This first release focuses on cinematography, defining terms across four categories, Lens, Shot Size, Camera Movement, and Camera Angle, with future versions expanding into composition, framing, depth of field, and creative areas beyond cinematography.

This Vocabulary was developed in the MovieLabs Industry Forum through collaboration among working creative professionals, subject matter experts, generative AI technologists, and tool developers. That mix is deliberate: the terminology has to reflect how productions actually talk and what AI systems actually need to be told.

The MovieLabs Creative Vocabulary formalizes the language film and TV industry have always used and makes it usable by the AI systems now entering their workflows. Additionally it also serves as a useful reference for everyone. For the creative community, that means their established language keeps working as a way to guide the AI systems, with no need to learn a separate machine dialect. For AI builders, it provides a precise target: model and system providers can tune their products to recognize this language, and AI agents can learn not just what each term means, but also the context for why an artist would choose it.

This work falls under, and aligns, with one of the four strategic focus areas for MovieLabs on AI: Metadata & Context – Enabling AI to act with greater precision and creative alignment through rich contextual guidance.

Why now?

Generative AI and Assistive AI are becoming additional creative options in production workflows. Creative teams are guiding image, video, audio, and 3D model generation through prompts and structured inputs, and they’re discovering something frustrating: the precise, expressive language of filmmaking doesn’t always survive the trip to an AI model or through an agentic workflow. Ask one tool for a low angle and you get a heroic, towering subject; ask another for the same thing and you may get something else entirely. The words are interpreted differently by each AI system and sometimes by the same AI system on subsequent passes.

The MovieLabs Creative Vocabulary exists to close that gap. Rather than inventing a new way to talk to machines, it preserves the language artists already know and makes it usable, consistently, across human collaboration and AI-assisted creation.

A key design choice sets the Vocabulary apart from traditional film-craft references: the Vocabulary emphasizes creative intent, not just technique. It doesn’t merely tell you that a Close-Up frames a character’s face; it tells you the Close-Up exists to intensify the drama by drawing the viewer into intimacy with a character and their emotional state. For a prompt driven AI model or an agent working on a storyboard, that intent describes the outcome the creative team desires.

For the creative community and AI assisted workflows

For creators working with generative tools, the Creative Vocabulary is a reference: a clear, concise guide to the terms, their definitions, and the creative purpose behind each one. It’s also an accessible on-ramp for emerging creators, film students and others who are fluent in AI tools but new to the language of filmmaking.

For AI agents assisting with storyboarding and pre-visualization

Because each term carries both a definition and its creative intent, the Vocabulary becomes something an AI agent can reason with, particularly the agents beginning to assist with storyboarding and pre-visualization.

Consider what it takes to turn a script into a shot list. A scene calls for a character to receive devastating news. How should that moment be captured? An agent that understands creative intent can begin to reason it through: the goal is to convey the character’s emotional state deeply and intently, so a Close-Up, dominated by the character’s face, draws the viewer into that emotional state, and an Extreme Close-Up on the eyes can push further still, isolating the smallest flicker of feeling. Elsewhere in the same script, a sweeping establishing moment might call for an Extreme Wide Shot to set scale and mood, or a Crane camera move to deliver an epic reveal.

This is the difference between an agent that is provided with what a shot is and one with both what it is and why to use it. By encoding the common reasoning that filmmakers carry in their heads, the Vocabulary gives storyboarding and pre-viz agents a jumpstart to translate narrative intent into visual choices instead of guessing.

For AI model providers, aggregators, and search platforms

The Vocabulary is just as much a practical resource for the companies building the AI systems. Supporting it isn’t just good citizenship, it directly improves how well artists can control and guide the tools, which is increasingly what they’re choosing models on.

For generative model providers, across image, video, and 3D, the Vocabulary offers a structured map of the creative concepts users will reach for. It can help :

• Spot gaps in training data, surfacing creative concepts that a model doesn’t yet handle well because they’re underrepresented in it’s training datasets. For example, a model might support shot sizes confidently but stumble on camera movement simply because the data isn’t there.
• Improve labeling consistency, identifying where data annotation has been applied ambiguously or incorrectly, for instance, where Close-Up and Extreme Close-Up are conflated, leaving the model unable to distinguish between concepts that artists treat as distinct.
• Strengthen prompt interpretation, recognizing the synonyms, acronyms, and alternative names artists use for the same idea, Close-Up and CU, Bird’s Eye View and Overhead Shot, often without retraining the model at all.

For model aggregators and platforms, especially those offering model-specific prompt enhancement, the Vocabulary provides a stable, shared reference for normalizing creative intent across very different underlying models. An artist can express what they want once, in familiar language, and a platform can translate and expand that intent into each model’s preferred phrasing. The user never has to learn the quirks of different prompting styles for different models; they just describe the shot they want.

For semantic search and media-discovery providers, consistent terminology is the difference between finding the right frame and scrolling for an hour. When shots, frames, and sequences can be pulled from the asset management systems against a shared vocabulary, users can search by creative intent, every drone shot in the archive, every telephoto close-up, every dutch-angle moment, rather than fighting with inconsistent free-text tags or low-level visual features. Aligning search engines with the Vocabulary makes that discovery reliable at scale, and portable across productions, studios, and tools.

In every case, the theme is the same: production teams spend less time wrestling with tools and more time creating when the tools can consistently interpret the creative language.

Open by design

The MovieLabs Creative Vocabulary is released under the Creative Commons Attribution 4.0 International (CC-BY-4.0) license. It can be freely used, shared and adapted, with attribution. Our goal is broad adoption across studios, vendors, and creative communities, because a shared language only works when it’s genuinely shared.

What’s next?

Version 1 is an important milestone, but still a milestone. Future releases will expand into other cinematography categories and reach into new creative domains, and we’ll be adding reference examples to make each term even clearer. We’d love for the community, filmmakers, artists and AI builders alike, to put it to work and help shape where it goes.

Reach out to the MovieLabs Industry Forum for specific suggestions on the MovieLabs Creative Vocabulary or to learn more information about the ongoing AI activities or the larger 2030 Vision.

The MovieLabs Creative Vocabulary v1 is available now under CC-BY-4.0. To explore the full set of terms, definitions, and creative intent, visit MovieLabs Creative Vocabulary.