Sources and provenance
What Rivu records about researched claims and production assets.
Research traceability
Research-led productions preserve the source context used for material factual claims. Names, dates, scores, statistics, quotations, scientific claims, and events should be supported by the research artifact.
Narrative interpretation, hooks, transitions, and payoff can use editorial language as long as they do not contradict the verified factual spine.
Asset records
Every sourced or generated asset retains production metadata where available:
- provider
- original source URL
- license or usage context
- acquisition or generation cost
- the scene and composition revision that used it
Provider names may appear in technical provenance views rather than throughout the main production experience.
Current provider layer
The current provider layer can work with stock and generated sources including Pexels, Pixabay, Unsplash, ElevenLabs, GPT Image, OpenRouter, and DeepSeek. Availability depends on project configuration and provider credentials.
Public API status
Rivu does not expose a public customer API during the private preview. Internal worker endpoints remain part of the application deployment. api.rivu.video is reserved for a future versioned public interface.