Veo 3 moved sound inside the generation workflow

What Google announced

On 20 May 2025, Google introduced Veo 3 with generated audio, including environmental sound and dialogue. The announcement also introduced Flow as an AI filmmaking environment bringing together Veo, Imagen and Gemini. These are company-described capabilities. The source is strong evidence for what was announced and when, but it is not an independent production benchmark.

A different unit of work

For a vertical-drama producer, the relevant question is not simply whether a model can make a talking character. Our workflow analysis asks whether the resulting shot can survive editorial use: dialogue timing, intelligibility, continuity with adjoining shots, performance direction and the ability to revise one part without breaking another.

Generating sound and image together can change where those decisions are made. It does not demonstrate that a separate sound pass, edit or quality-control step becomes unnecessary. A successful promotional clip is one selected result; it is not a record of the attempts, labour or finishing required to reach it. We have not tested the model for this article.

Measure the accepted output

A useful production experiment would keep a small log: the brief, number of attempts, accepted seconds, operator time, finishing work and reasons for rejection. Those fields let a team distinguish an attractive example from a repeatable process. They also make comparisons possible without treating a vendor's showcase as representative of every output.

The cost question should then attach to the completed, usable scene. A generation charge alone cannot account for discarded attempts or downstream labour. This is an accounting framework, not a claimed saving or a forecast of job replacement. It remains useful whether a team uses generated material for a storyboard, a pitch, a background shot or a released sequence.

Keep launch facts separate from today's product

The inherited draft combined a 2025 announcement with later model specifications and live pricing. This edition deliberately narrows the record to the launch and its workflow implications. It makes no current price, maximum-duration or availability promise. Google's original announcement collage is retained below and explicitly labelled as synthetic promotional imagery supplied by Google, not documentary photography or artwork generated for Vertical Ledger.