CampaignReel

Ad placement specs against what generators output

Which entries reach a minute in a single call

A sixty second non-skippable is the format that splits this register. Three entries reach it in one call: thirty minutes per scene, ten minutes, and two minutes in six second increments. The rest assemble it from between two and ten pieces. As of 2026-09-22.

Three routes to sixty seconds, no joinTwo entries animate a supplied audio track, so the length follows the speech. The third cuts a two minute request into six second beats itself.Audio driven30 minutes per sceneSpeech sets the lengthNo seamsAudio driven10 minutesSpeech sets the lengthNo seamsBeat drivenTwo minutesThe model cutsSeams nobody placedA sixty second slot in one requestWhat each route gives up
Fig. 1 These are not simply the most capable rows; two of them are a different product answering the same brief.
The three entries that clear a minute without a join. Recorded 2026-09-22.
ItemWhat is publishedDetail
HeyGen30 minutes per sceneA performance driven by supplied audio, fifty scenes in one job
Hedra10 minutesA performance driven by supplied audio, no assembly figure stated
Amazon Nova ReelTwo minutesSix second beats inside one request, on a particular model identifier

Inclusion rule. Entries whose own page states a length of sixty seconds or more for a single request. A length reached by assembling several requests is not counted here. Order. Longest published figure first.

1Two of the three are not generating a scene

Where the picture follows a supplied audio track, the output runs as long as the track. The ceiling is a statement about file handling rather than about how long a generated scene holds together, which is what the same number means on a prompt-driven row.

So the three are not simply the most capable rows in the register. Two are a different product answering the same brief, and choosing them is a decision about whether the advertisement is a person speaking.

2The third makes the cuts itself

Two minutes in six second increments inside one request means the joins exist and nobody in the edit placed them. That removes the seam-matching work and removes the ability to decide where a beat lands.

For an advertisement with a fixed beat at the end that is a real loss, and it is not visible in the length column. It also carries a condition: anything past the first six seconds needs a particular model identifier.

3What the other entries pay for the same minute

Two calls on the widest prompt-driven ceiling here, four on the several entries that stop at fifteen seconds, eight where the ceiling is eight, and ten where the length is fixed at six. Every extra call is a seam.

Impression billing sharpens that. A non-skippable delivers and charges for its whole length, so a visible join is a second the advertiser paid to show, and the join count is the number this comparison actually decides.

4A generator's ad pipeline

For ad work around “A minute in one call”, SceneMixer's pages explain how its ad projects are produced:

Measured across every job that finished successfully on scenemixer.com in the thirty days to 20 September 2026: analysis of a whole book runs about 2 minutes 49 seconds at the median, one episode's storyboard 1 minute 54 seconds, one video segment 3 minutes 20 seconds, and compositing an episode 3 minutes 52 seconds.

SceneMixer, Turn a Novel into a Video Series with AI — “How long does it take to produce one episode?” (read 2026-10-13)

A serialized vertical-video show (typically 60–120 seconds per episode, 9:16) where the cast, scenes, and footage are AI-generated.

SceneMixer, The AI Micro Drama Production Pipeline — “What is an AI micro drama?” (read 2026-10-13)

5Sources

Format requirements come from Google Ads Help, read 2026-09-12, and tool limits from each vendor page listed on published limits, read 2026-09-22. The rule against mixing the two is on method. Nearby: Ceilings on inputs, Sound that arrives, Wider than widescreen.