Video Voiceovers in ImageLayer: Current Launch Status

Product status — August 25, 2026: In-widget merged-video voiceover is not part of the current production launch path. Do not promise that a signed-in organization can create or download a merged narrated MP4 until the workflow is explicitly enabled and verified for that organization.

ImageLayer retains a capability-gated merged-video voiceover prototype in the codebase, but the production configuration does not expose its action and keeps the workflow fail-closed while its provider operation and financial settlement are hardened. Base video generation and standalone Audio mode are separate requests with separate availability and quotes.

Launch-safe workflow

1. Generate and approve the base clip in Video mode.
2. Write a narration script that fits the clip when read aloud at a normal pace.
3. Generate the narration as a separate request in Audio mode, using a currently available public TTS model and compatible voice.
4. Review the exact quote shown for each request; do not assume one fixed voiceover credit price.
5. Listen to the complete audio file and check its duration against the video before editing them together outside ImageLayer.
6. Review the complete final export. Speech must finish before the video ends, with no clipped word, abrupt cutoff, or unsupported spoken claim.

This workflow does not imply that the product currently returns a merged MP4. It describes a manual handoff that keeps the base video and standalone narration independently reviewable.

Why the old examples were removed

The former Prism and NovaCraft narrated MP4 examples ended while their speech tracks were still active. Both files were removed from the public site rather than presented as launch-ready output.

Prism and NovaCraft are fictional teaching brands. Any future replacement must disclose that status in the page metadata, visible copy, and narration brief.

Acceptance checklist for a future merged result

• The in-widget voiceover action is enabled for the organization and allowed by the current production configuration.
• The base video and TTS requests each have a current pre-submit quote and settled usage record.
• The exact audio file is bound to its complete transcript, model, voice, duration, and file hash.
• Human playback confirms every spoken word and the final silence.
• Speech does not continue past the final video frame.
• The merged file decodes correctly and has the intended video and audio streams.
• Fictional examples are labeled; factual and product claims have evidence.

No narrated MP4 or standalone MP3 is embedded in this article until a replacement meets all of those checks.

---

Keep exploring

• Text-to-Speech in ImageLayer: Current Models, Voices, and Quotes — Public TTS models, compatible voices, and per-request pricing.
• Fictional Prism Launch Kit — A clearly labeled multimodal concept workflow.
• Fictional NovaCraft Campaign — A clearly labeled brand-campaign concept.

Check current availability

Open the playground and sign in to see which Video and Audio workflows are enabled for your organization. The interface and request quote are the current source of truth; this article does not enable a feature.