We Built an AI Video Glossary. Here's Why.
AI video production moves fast. The tools change every few months, new roles are emerging that didn't exist two years ago, and half the terminology sounds like it was invented by someone who's never been on a film set.
That's a problem — because when clients, producers, and creative teams don't share a common language, projects get messy. Briefs get misunderstood. Budgets get questioned. And the gap between what AI can actually do and what people think it can do keeps getting wider.
So we built a glossary. Not a tech blog explainer or a Wikipedia deep-dive — a practical, no-jargon reference covering 87 terms that actually matter in AI video production right now.
What's in it
The glossary covers three areas in a single A–Z:
The tools and techniques — from diffusion models and LoRAs to ControlNet, ComfyUI, and the major platforms like Runway, Veo, and Midjourney. These are the building blocks of how AI video actually gets made, explained in plain language.
The business of AI production — this is the part nobody else is covering. New roles like AI Creative Director and AI Producer. New cost lines like Compute and Licensing. New contractual concepts like Craft Intelligence IP and Creative Governance. These terms are being developed in collaboration with the APA, and they matter because AI production introduces budget structures and scope questions that traditional frameworks weren't designed for.
The current model landscape — where things stand with Sora (shut down March 2026), Runway Gen-4, Veo's native 4K and audio generation, Kling's 120-second clips, and the open-source Wan model family. We'll keep these updated as the market moves.
What's next
We'll keep updating the glossary as the landscape shifts. Terms will be added, definitions will evolve, and some entries will probably become obsolete faster than we'd like.
We're also building a separate General Video Production glossary — covering the traditional vocabulary from pre-production through to delivery — for clients who are newer to the process.
If there's a term we've missed, or one you keep having to explain to clients, let us know. This is meant to be useful, not comprehensive for the sake of it.
