VideoGenReview

What each AI video model's release notes say changed

Four different numbers called clip length

Notes use length in four ways. The first is a single generation: Hailuo 02 up to 10 seconds, Wan 3.0 up to 30. The second is a clip grown by extension: Gen-3 Alpha up to 40 seconds in total. The third is a take driven by a reference: Kling 2.6 Motion Control, 30 seconds of action in one take. From 5 source pages, read 2026-09-25; no maker pays to be listed.

Where this note's examples come fromThe examples in this note come from 5 entries across 4 model lines.ExamplesKling AI2 entriesMiniMax Hailuo1 entryRunway1 entryWan1 entry
Fig. 1 Each example links to its line page.

The fourth appears in avatar and lip-sync notes, which can run far longer, five minutes for Kling's avatar videos, because the picture is driven by a track rather than invented.

The entries this note draws on, read 2026-09-25.
DateLineVersionWhat changed
4 September 2024RunwayGen-3 AlphaGen-3 Alpha clips can be extended to 40 seconds in total
18 June 2025MiniMax HailuoHailuo 02Hailuo 02 generates 1080p and clips of up to 10 seconds
4 December 2025Kling AIKling AvatarAvatar videos can run five minutes
17 December 2025Kling AIKling 2.6Kling 2.6 Motion Control copies an action and expression, up to 30 seconds in one take
6 August 2026WanWan 3.0Wan 3.0 folds reference, editing, replication and driving into one model, with clips up to 30 seconds

Inclusion rule. Entries used as examples in this note. Order. By date, oldest first; undated rows last.

1In the notes

Gen-3 Alpha videos can be extended an additional 5 or 10 seconds to create up to 40s of generated video.

Runway, changelog

Released MiniMax Hailuo 02, a next-generation video generation model with support for 1080p resolution and video generation of up to 10 seconds.

MiniMax, model release notes

Ability upgrade, supporting the generation of 5-minute avatar videos

Kling AI, API updates

30 seconds of high-difficulty action in one seamless take, with total command over movement, expressions, lip-sync, and gestures.

Kling AI, release notes

It supports four-modal all-reference input, up to 30 seconds of video generation, and can parse files, web pages, and complex images.

Alibaba Cloud Model Studio, newly released models
  • Extensions in Gen-3 Alpha
    Gen-3 Alpha clips can be extended to 40 seconds in totalRunway, changelog / since 2024-09-04 / checked 2026-09-25
  • MiniMax-Hailuo-02
    Hailuo 02 generates 1080p and clips of up to 10 secondsMiniMax, model release notes / since 2025-06-18 / checked 2026-09-25
  • 12/04/2025
    Avatar videos can run five minutesKling AI, API updates / since 2025-12-04 / checked 2026-09-25
  • 2025-12-17
    Kling 2.6 Motion Control copies an action and expression, up to 30 seconds in one takeKling AI, release notes / since 2025-12-17 / checked 2026-09-25
  • wan3.0-video
    Wan 3.0 folds reference, editing, replication and driving into one model, with clips up to 30 secondsAlibaba Cloud Model Studio, newly released models / since 2026-08-06 / checked 2026-09-25

2A pipeline on current models

Release notes such as “Four different numbers called clip length” matter once a pipeline runs the models; SceneMixer's guides describe how it does:

Paste a scene or a short script and SceneMixer gives you a storyboard — numbered shots with framing and seconds — and then a moving version of each segment, which you can join into a rough cut of the scene.

SceneMixer, Using SceneMixer for Previs and Animatics Before a Live-Action Shoot (read 2026-10-13)

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)

3Other reading notes

4Notes read