Model family

The Kling VIDEO versions and what separates them

The Kling 3.0 series is not one model, it is two. Their own guide says VIDEO 2.6 was upgraded to VIDEO 3.0 and VIDEO O1 was upgraded to VIDEO 3.0 Omni, so there are two parallel lines each with its own successor.

  • two parallel lines
  • fifteen seconds from 3.0
  • five languages, all new

current Maker: Kuaishou

Last checked: This is a reference page. It is re-checked against the vendor sources and updated when a new version ships. What changed

The version to use

3.0, and 3.0 Omni if your work involves multimodal input. There is no reason to stay on 2.6 today unless a service has been tested on it, because everything 3.0 adds sits on top of what 2.6 already did.

Something only their own table gives you

The official VIDEO 3.0 guide carries a capability table comparing 2.6 and 3.0 column by column, and what it shows is sharper than any summary. Text-to-video, image-to-video, start and end frames, and native audio: all four are present in both versions. So audio is not something 3.0 introduced.

What 3.0 did introduce is this: multi-shot in a single generation, element reference on the start frame, multi-character coreference beyond three, multilingual support, dialects and accents, fifteen second output and flexible duration. All seven are marked absent in the 2.6 column.

Five languages that are younger than you would think

The Kling speech language list, Chinese, English, Japanese, Korean and Spanish, is in that same table a new capability of 3.0. The multilingual row is marked absent for 2.6. So until February 2026 this model had no language list at all.

For a Persian reader that has two consequences. Persian is not among the five, so Kling produces no native Persian speech. And because the list is new, it may well grow; this sentence carries a date and that date is 13 August 2026.

4K, and which version it is attached to

Kling says it launched native 4K on 23 April on the 3.0 series. So the capability is bound to the whole series rather than to one of the two lines. That single number is what puts Kling ahead of Seedance in the resolution column of our table, while Seedance still wins overall.

Specifications

Every number here comes from the vendor's own page, with that page linked beside it.

Clip length in one pass
15 seconds Source
Video resolution
4K Source
Vertical lines
2,160 lines of vertical resolution Source
Synchronised audio
yes Source

Release timeline

Each version with its own release date, and what changed against the one before it.

  1. Kling VIDEO 3.0 current Against VIDEO 2.6: multi-shot, element reference, multi-character coreference, five speech languages and the 15 second duration, none of which 2.6 had.

Where it falls short

  • Persian is not on the five-language speech list, and that list itself only arrived with 3.0.
  • Kling is absent from the Runway price list, so its per-second cost cannot be compared with the other candidates.
  • We do not have the full version timeline, because the Kling release-history page renders in JavaScript.
  • No supported-countries list is published, so the Iran column for this family reads unknown.

Our take

Go to 3.0, and take Omni if you have multimodal input. If you want Persian output with speech, this family does not answer yet and you should budget for a separate voice step.

Sources

  1. Kling AI, VIDEO 3.0 model user guidevendor sourcekling.airead on 13 August 2026
  2. Kling AI, VIDEO 3.0 Omni model user guidevendor sourcekling.airead on 13 August 2026
  3. Kling AI, native 4K video model announcementvendor sourcekling.airead on 13 August 2026