custom voiceover feature

The request
When generating video in Firefly (image-to-video, including partner models like Kling), let users upload their own audio recording and have the character’s mouth lip-synced to that performance.

What exists today
As far as I can tell, Firefly can generate dialogue from the text prompt, but only in the model’s own voice. Firefly’s lip sync is limited to Translate Video and enterprise plans. There’s no way to bring your own recorded voice acting and have a generated character perform it.

Why it matters
Many creators cast their own voice talent: podcast hosts, filmmakers, animators, and brands with a signature voice. Right now, to get a character to speak our recorded lines, we have to leave Firefly for a separate lip-sync service. That breaks the workflow, splits credits across tools, and moves the footage out of Firefly’s commercial-safe pipeline.

My use case
I’m producing a short horror film where AI-generated characters are voiced by the real hosts. I build still frames, animate them with image-to-video in Firefly, and need the on-screen mouths to match the recorded lines.

How it could work

• An “Add audio” slot next to the Frames slot, with trimming to the clip length
• Lip sync to the uploaded audio, working on both stylized and realistic characters
• An option to keep my audio as the final track, or generate sound effects around it
• Support for the partner models, with the credit cost shown before generating
• In the Firefly video editor: select a generated clip and an audio clip on the timeline, then choose “Lip sync to audio”

Why this fits Firefly
It keeps performance control with the creator and works well with consistent characters, which fits your focus on commercially safe, production-ready video.