VideoGenReview

What each AI video model's release notes say changed

How to read a model line's page

A line page lists every entry for one maker's models, oldest first, with four kinds of source: the maker's own note, a platform's arrival note, an undated row from a model list, and an undated row from a platform's tier list. The Source column says which. From 3 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 3 entries across 3 model lines.ExamplesKling AIMidjourney videoSeedanceEvery layer: 1 entry
Fig. 1 Each example links to its line page.

The count at the top reflects how the maker writes notes. Kling's thirty-six entries come from a detailed API log; Midjourney's four from a blog. Read the versions list to see the steps, and the kinds of change to see where the notes spend their words.

The timeline at the top of the entry table shows the rhythm. Some lines post in bursts around a launch; others post a little every few weeks. A long quiet stretch on a line with a detailed log is worth noticing; on a line that only posts launches, it is normal.

Each version with three or more entries has its own page, and each kind of change with three or more entries for the line links to a page for that pair; the two PixVerse pages that dropped to two entries on the 2026-10-25 re-read stay up.

The entries this note draws on, read 2026-09-25.
DateLineVersionWhat changed
19 June 2025Kling AIKling 2.1Kling 2.1 image-to-video in standard and pro modes, plus a 2.1 Master
25 July 2025Midjourney videoMidjourney V1 videoLooping, set start and end images, and video in the Discord bot
Undated version rowSeedanceSeedance 2.5Seedance 2.5: 4 to 30 seconds, 1080p in 10-bit

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

1In the notes

Update image2video kling-v2-1 model, supports STD mode and PRO mode

Kling AI, API updates

You can now make videos with a specific start and end images

Midjourney, looping and end frame announcement

Duration: 4–30 s

BytePlus ModelArk, model list

2Using the models, not reading about them

Release notes such as “How to read a model line's page” matter once a pipeline runs the models; SceneMixer's guides describe how it does:

Qwen-Image 3.0 is the tier every account starts on for character sheets, location images and prop images.

SceneMixer, Qwen-Image 3.0: the default image tier (read 2026-10-13)

Each scene gets an automatic AI 3D model, and every shot in the scene is generated against it.

SceneMixer, Why a Whole Series Stays Consistent on SceneMixer — “Does the layout of a room really stay the same across shots?” (read 2026-10-13)

3Other reading notes

4Notes read