What AVS is
Agentic Video Studio is a Next.js application with one active product surface in frontend-beta/.
The current beta combines:
- A project dashboard for creating, importing, duplicating, and organizing work
- A browser-based editor with timeline, transcript, preview, assets, chat, and tool panels
- AI generation for video, image, music, speech, and prompt-driven project creation
- Shared video links, embeddable review pages, and export workflows
- Workspace, billing, BYOK, token, and storage administration
- A companion CLI flow for native recording import
- A public HTTP API for async video generation jobs
Main product surfaces
Marketing and public pages
The beta app ships public routes for:
/landing page/about/demo/docs/cli/status/privacy/tos
These pages explain the product, installation path, documentation, and operational status.
Auth and onboarding
The auth surface includes:
- Email/password login
- Forgot-password flow
- Email verification flow on
/auth/action - Onboarding completion flow on
/auth/onboarding - Magic-link generation support on the backend
Lifecycle emails and verification emails are app-owned and sent through the current messaging layer when configured.
Projects app
The authenticated app shell under /app includes:
- Project dashboard
- Shared links manager
- Settings
- Editor entry points
This is the operational center for cloud and local projects.
Editor
The editor is the product’s deepest surface. It includes:
- Desktop and mobile layouts
- Assets, preview, timeline, settings, chat, transcript, and toolbox panels
- Branching, merge, and branch comparison
- Presence, active-editor indicators, and cursor collaboration
- Command menu, shortcuts modal, tutorial flow, and notifications
- Local and cloud project modes
- Local and cloud rendering plus share-link creation
See Editor Overview for the panel-by-panel breakdown.
Machine interfaces
AVS exposes two major machine-facing entry points:
- Vidova CLI for native recording, packaging, auth, and import
- Public Video API for prompt-to-video and JSON-to-video async jobs
Core workflow loops
1. Project-first editing
Create a project from the dashboard, open it in the editor, upload or generate assets, then arrange clips on the timeline and export.
2. Prompt-first creation
Use the project-creator flow to start from a text prompt, optional asset uploads, and a selected model. AVS creates the project first, then continues the generation flow in the editor context.
3. Asset-first editing
Import existing footage, transcripts, audio, images, screen recordings, 3D models, or custom components, then assemble them with timeline tools and inspector controls.
4. Machine-assisted creation
Use chat in Ask, Agent, or Plan mode, or run specific tools from the toolbox to manipulate assets, clips, layers, and generation jobs.
5. Delivery and review
Render locally or via the backend, then download, create a public or org-only shared link, or embed the shared player elsewhere.
Product capabilities at a glance
Creative primitives
- Video, audio, image, text, component, 3D model, blur, vidova, and zoom clip workflows
- Scene configuration for name, dimensions, frame rate, and background color
- Text templates, component templates, and custom Motion Canvas components
- Video effects jobs and chroma-key workflows
AI primitives
- Veo video generation
- Gemini image generation
- Lyria music generation
- TTS generation
- Prompt expansion and prompt-to-project creation
- Chat attachment uploads and model selection
Collaboration and operations
- Branches and merges
- Workspace membership and invites
- Shared links
- Local/offline project support
- BYOK provider configuration
- Public API tokens and CLI tokens
Recommended reading
- Start with Getting Started if you are new to the beta.
- Use Feature Reference for a full inventory.
- Use Projects & Workspaces for storage and collaboration structure.
- Use Editor Overview for the editing surface.
