How Veodra works
Not a generator you prompt — a production pipeline you brief.
The AI Director
The Director is the part of Veodra that makes decisions a human producer would otherwise make: how many scenes a brief needs, which provider fits each one, what pacing and voice tone the piece calls for, and when a scene needs to be re-tried. You never talk to Veo or Kling directly — you talk to the Director, and it talks to them.
A project is a graph, not a folder of files
Every video is modelled as a project tree, not flat exported files — so “re-do scene 14” means regenerating one node, not starting over.
Know the cost before you commit
Before rendering starts, the system estimates the full cost from the per-provider price list and can suggest a cheaper mix with the same scene count. This is an illustrative example from our published concept notes, not a live quote:
- 6-minute video, 24 scenes (12 Kling · 6 Veo · 6 Runway) — video ≈ $7.4, voice ≈ $0.5, music ≈ $0.2, storage ≈ $0.1 → total ≈ $8.2
- Swap Veo for Kling on background scenes and the same brief renders for meaningfully less, with a minor quality trade-off the system flags upfront.
What's under the hood
Veodra doesn't build its own video model — it orchestrates the best ones available and keeps the choice invisible to you. Planned providers, per the published concept:
Questions people ask first
Before you ask us to build your video, here's what we can already answer.
Is this live yet?
No. Veodra is in development — the concept and data model are defined, but there's no self-serve signup or live rendering yet. This site exists to collect early-access interest.
Do I need to know which AI model to use?
No — that's the whole point. You describe the outcome and pick a quality tier; the Director chooses Veo, Runway, Kling or another model per scene.
Can I redo just one scene?
That's the design goal: a project is a graph of scenes, not a flat export, so a single scene can be regenerated without re-rendering the whole video.
How is this different from Runway or Kling directly?
Those are generators — you write prompts and assemble the result yourself. Veodra is the layer above them: it plans the scenes, picks the provider per scene, adds voice and music, and delivers a finished cut.
Is Veodra part of Loquify or web3e?
It's a separate product in the same holding, sharing infrastructure (identity, crypto payments) with web3e — see the About page.