How long a generated shot with speech can run

Six entries publish a length. Two give a menu of fixed shot lengths, two give a range to choose from, one gives a ceiling in minutes, and one publishes a free allowance. The rest say nothing. From 3 source pages, read 2026-09-22; no maker pays to be listed.

Inclusion rule. Entries whose documentation states a duration for a generation, its audio, or the allowance a plan includes. A duration mentioned in marketing material without appearing in the documentation read for this register does not earn a row. Order. Alphabetical by model name.

1A menu, a range and a ceiling are three different constraints

A fixed menu of four, six or eight seconds forces a scene to be written in those units. A range of two to thirty leaves the decision to a writer. A ceiling of ten minutes is not really a constraint at all for drama, because nothing holds a single generated shot that long.

Reading them as one number misses the difference. The menu is the one that reaches a script: dialogue written in eight-second beats survives the model, and dialogue written to a thirty-second paragraph has to be cut before it can be generated.

2The longest published figure belongs to a talking head

Ten minutes of supplied audio for a talk is the largest number here, and it comes from a product aimed at presentations rather than drama. A single unbroken take is exactly what a corporate video wants and exactly what a short drama never uses.

So the length column is not a ranking. A short ceiling on a generator that cuts every few seconds is a better fit for a series than a long one on a product that expects a single speaker facing the camera.

3Extensions turn a ceiling into a sequence of joins

One entry reaches two minutes by adding up to twenty seconds at a time. Each addition is a join, and a join is a place where a voice, a level or a room tone can shift. The published total is therefore the length of an assembly rather than of a take.

Treated as a budget for one beat rather than one scene, the same figure becomes useful. Treated as a promise of two continuous minutes it will disappoint, and nothing in the documentation says which reading is intended.

• Shot length a line has to fit
  Generations run 16 or 20 seconds, and an extension adds up to 20 seconds at a time to a total of 120OpenAI, video generation guide / recorded 2026-09-22
• Audio source
  A script is either text of up to forty thousand characters or an audio url, with audio limited to five minutes for clips and ten for talksa script read aloud or a fileD-ID, create a talk reference / recorded 2026-09-22
• Length per call on the free tier
  A free account runs one generation a month with a fifteen-second ceiling, with paid limits following the planSync, sync-3 model documentation / recorded 2026-09-22

4Speech, generated with the picture

For the speech side of “How long a shot can be”, SceneMixer's guides describe how lines are spoken:

Dubbing replaces the audio of a finished clip; the picture was generated for a different line, so mouth shapes, pauses and shot length all belong to the original language. Native-language generation gives the video model the line in the target language before it renders the shot, so the performance, the pacing and the cut points are built for that language.

SceneMixer, Native-Language Dialogue vs. Dubbing in AI Video (2026): Why the Cast Should Speak the Line (read 2026-10-13)

Scene order, dialogue and beats follow the script; what the pipeline adds is coverage — which shot sizes, who is in frame, how long each shot holds.

SceneMixer, Script to Video, End to End — “How faithful is the shot list to my script?” (read 2026-10-13)

5Sources

Every line is read from the vendor documentation linked on the speech controls page, checked 2026-09-12. What counts as documented is on how read. Nearby: Fifteen seconds of voice, Answers software can check, What silence costs.