Volarra brings planning, storyboarding, generation, and timeline editing into one project.
This is a plain-language guide to the things it works with — and how they fit together.
Everything starts from a script. You develop it by talking to an AI
Director in the Story Conference, which returns a structured
Project Plan — a breakdown into Scenes and Shots.
Each shot is one unit of generation: plan it with a Storyboard Frame, then
generate its Video. The version you pick becomes a Clip on the
Timeline, where clips lay end-to-end across Tracks and render to
an exported video. Running across all of it are reusable Creative Entities that
keep things consistent — plus full provenance for every generation.
How it fits together
From a script to a finished video.
The four workflows usually split across separate tools. In Volarra they share one project.Named concepts and how they relate. The Shot is the hub — almost everything is generated for a shot.
Where you work
The surfaces.
These are the windows you spend time in — places, not things you make.
Project Chooser
The top-level entry point. Open a recent project or start a new one. One project is open at a time.
Director
The planning center. Hold the Story Conference, edit the script, scenes, shots, prompts, and casting, and generate storyboard frames.
Storyboard
A whole-film visual review: every shot's frame at once, in order, so you can judge pacing and continuity at a glance.
Editor
The timeline-based assembly surface: a program-monitor preview, multi-track timeline, version picking, and export.
Entity Studio
A global workshop for creating and refining Creative Entities outside of any single project.
Generation Queue
One window showing every in-flight and recent generation across the app, with live status, cost, and retry / cancel.
Glossary
Every term, defined.
The project & the script
Project
A single video you're making, and the durable container for everything in it — the plan, the generated media, the conversation history, and the export settings. On disk it's a plain folder of standard media files and readable metadata, so you can inspect, back up, or hand it off without Volarra in the loop.
Script
The creative through-line of the project: what is said and what happens. Volarra treats the script as the most important artifact — every scene, shot, frame, and clip downstream is an interpretation of it. You develop it in the Story Conference and can also edit any part by hand.
Project Plan
The structured form of the script — broken into ordered Scenes and Shots, with project-level settings like title, aspect ratio, and target length. It's the blueprint the timeline is built from, not yet a full editing timeline.
Story Conference
The ongoing, plain-language conversation with the AI Director that writes and revises the script. You give notes like a producer would; each turn returns updated, directly editable structure — not prose — and summarizes exactly what it changed. The conversation is saved as the script's revision history.
Director
The AI-assisted planner behind the Story Conference. It turns a plain-language brief into a structured Project Plan and revises it turn by turn. It's provider-neutral — it can run on different AI models behind one consistent contract.
Story structure
Scene
A narrative unit that groups one or more Shots. Carries its own title, synopsis, target duration, and any Creative Entities (a recurring location or character) that apply to every shot inside it.
Shot
The core unit of generation — one planned piece of video or audio within a scene. A shot holds its prompt, modality, camera and audio notes, target duration, script lines, the entities it casts, and its storyboard frame(s) and video version(s). Almost everything you generate, you generate for a shot.
Script line
One spoken line belonging to a shot — either narration or character dialogue. Script lines are part of the shot's generation inputs and are how the script's words attach to specific moments.
Modality
What kind of generation a shot uses — text-to-video, image-to-video, reference-to-video, text-to-speech, sound effect, music, or imported media. The modality determines which providers and models can fulfill the shot.
Visual planning
Storyboard frame
A still image representing a shot's intended look, generated before you spend time and budget on video — from a quick sketch to a high-fidelity keyframe. A selected frame can also become a reference that seeds the shot's video generation.
Reference
An input that steers a generation toward a desired result — most often an image. A shot's references can include its selected storyboard frame and the canonical images of the Creative Entities cast into it, so the result stays on-model.
Generated media & versions
Generated video
The actual video a provider returns for a shot. A shot can accumulate several over time — each generation is kept as a retained version, never silently overwritten, so your iteration history is preserved.
Version / variant
One attempt at a shot's media. Because AI generation produces many near-misses, Volarra keeps every attempt and lets you pick which one is active; the active version is what flows to the timeline.
Clip
A piece of media placed on the timeline — the timeline-level instance of a shot's chosen video, positioned on a track with a length. The distinction worth remembering: a generated video is the file; a clip is that file placed in your edit.
Prompt
The text instruction for a generation. In Volarra each shot supplies its own prompt.
Prompt template
A reusable wrapper of system guidance and structure, authored once at the app level. Your shot prompt fills into the selected template — you don't author a template per shot.
Resolved prompt
The final text actually sent to the provider — your shot prompt combined with the template, the cast entities' guidance, and the shot's script. It's always inspectable before generation, so what you see is what gets sent.
Provider
An external service that does generation work — for the Director, or for images / video / audio. Volarra is provider-agnostic: connect several, compare them, and avoid lock-in.
Model
A specific generative model offered by a provider. The same modality can be served by different models that differ in quality, cost, speed, and aspect ratio — Volarra keeps those differences visible.
Generation record
The durable, inspectable record of a single generation attempt — provider, model, resolved prompt, inputs, parameters, status, cost, and output. This is what makes generated media reproducible and auditable.
Generation queue
The single place all generation work runs — plans, frames, and videos — across every window. It runs many jobs concurrently (bounded per provider), shows live status and cost, backs off on rate limits, and lets you cancel or retry. It runs on your machine against your own provider accounts, not a Volarra cloud.
Provenance & cost
The metadata Volarra preserves for everything it generates — provider, model, prompt, settings, seed, timestamps, rights, and cost. Cost is tracked as an estimate before you generate and actual spend after, including cumulative cost per shot.
The timeline
Timeline
The arrangement of your clips over time in the Editor — the assembly and playback model for the finished video. Today it's a strictly linear sequence: clips lay back-to-back in shot order. Playback and export use the same interpretation.
Track
A horizontal lane holding one type of media — video, narration, music, or sound effects — each with its own volume and mute / solo controls.
Playhead
The marker indicating the current time position during playback and scrubbing.
Export
Rendering the whole timeline into a single standard video file (MP4). Preview and export composite the timeline the same way, so the export matches what you previewed.
Format set Roadmap
A predefined group of output formats — 16:9, 9:16, 1:1, 4:5 — that generate and export every aspect-ratio variant in one pass. Built for advertising and paid-media work.
Reusable creative assets
Creative Entity
A persistent, reusable subject that can appear across many projects — a character, prop, location, vehicle, visual style, voice, sound palette, or music motif. It bundles a description, identity profile, canonical reference set, variants, generation guidance, and rights metadata. Entities are how Volarra fights drift, keeping a character or look consistent from shot to shot and project to project.
Media asset
A concrete media file — image, video, audio, voice sample, music cue, sound effect, source video, or style reference — together with its description, tags, and provenance. Generated videos and storyboard frames are media assets attached to a shot.
Canonical reference
The preferred reference media for a Creative Entity — the definitive “this is what it looks/sounds like” set that gets fed into generation to keep the entity on-model.
Library
A searchable collection of Media Assets and Creative Entities. The Global Library is reusable across every project; a project's pool is what's been brought into or invented for one project. Entities live in the global library as portable folders, independent of any project.
Entity Studio
The global workshop for authoring Creative Entities outside any project — describe one, generate candidate directions, iterate, generate variants, mark canonical references, and write its profile.
Brand Kit Roadmap
A curated, shareable set of entities, assets, style references, and default prompt guidance that a team reuses across projects to keep campaign work consistent. A portable folder any project can reference — a consistency tool, not an access-control mechanism.
Workspace & ownership
Local-first project folder
Volarra stores each project as an ordinary folder of standard media files and human-readable metadata on your own disk. Nothing is trapped in a proprietary cloud — open, move, back up, or inspect a project without the app.
Shared workspace Roadmap
A single folder path that several Volarra installations point at — a network share or mounted cloud bucket — to collaborate without any Volarra cloud infrastructure. Concurrent access is handled with file-level locking and surfaced conflicts.